title
Image Processing Tutorial Using Python | Python OpenCV Tutorial | Python Training | Edureka
description
๐ฅ Python Developer Masters Program (๐๐ฌ๐ ๐๐จ๐๐: ๐๐๐๐๐๐๐๐๐): https://www.edureka.co/masters-program/python-developer-training
This Edureka Live video on "๐๐ฆ๐๐ ๐ ๐๐ซ๐จ๐๐๐ฌ๐ฌ๐ข๐ง๐ ๐๐ฎ๐ญ๐จ๐ซ๐ข๐๐ฅ ๐๐ฌ๐ข๐ง๐ ๐๐ฒ๐ญ๐ก๐จ๐ง" will provide you with a comprehensive and detailed knowledge of Image processing and how it can be implemented using OpenCV library. In this video, you will be working on Image processing with Python and also create a model using a convolutional neural network. Finally, we will build an end-to-end model to process and identify the handwritten images. These are the following topics that are covered in this video on Image Processing Tutorial Using Python :
00:00:00 Introduction
00:00:52 What Is Image Processing?
00:02:40 Python For Image Processing
00:03:20 Image Processing Concepts
00:08:53 Digit Recognition Board
๐นEdureka Python Tutorial Playlist: https://goo.gl/WsBpKe
๐นEdureka Python Tutorial Blog Series: http://bit.ly/2sqmP4s
๐ดDo subscribe to our channel and hit the bell icon to never miss an update from us in the future: https://goo.gl/6ohpTV
๐๐๐๐ฅ๐๐ ๐ซ๐๐ฆ: https://t.me/edurekaupdates
๐๐๐ฐ๐ข๐ญ๐ญ๐๐ซ: https://twitter.com/edurekain
๐๐๐ข๐ง๐ค๐๐๐๐ง: https://www.linkedin.com/company/edureka
๐๐๐ง๐ฌ๐ญ๐๐ ๐ซ๐๐ฆ: https://www.instagram.com/edureka_learning/
๐๐
๐๐๐๐๐จ๐จ๐ค: https://www.facebook.com/edurekaIN/
๐๐๐ฅ๐ข๐๐๐๐ก๐๐ซ๐: https://www.slideshare.net/EdurekaIN
๐๐๐๐ฌ๐ญ๐๐จ๐ฑ: https://castbox.fm/networks/505?country=IN
๐๐๐๐๐ญ๐ฎ๐ฉ: https://www.meetup.com/edureka/
๐๐๐จ๐ฆ๐ฆ๐ฎ๐ง๐ข๐ญ๐ฒ: https://www.edureka.co/community/
#Edureka #PythonEdureka #PythonImageProcessing #ComputerVision #PythonProgramming #PythonTraining #PythonOpenCV #EdurekaTraining
---------๐๐๐ฎ๐ซ๐๐ค๐ ๐๐ง๐ฅ๐ข๐ง๐ ๐๐ซ๐๐ข๐ง๐ข๐ง๐ ๐๐ง๐ ๐๐๐ซ๐ญ๐ข๐๐ข๐๐๐ญ๐ข๐จ๐ง---------
๐ต DevOps Online Training: http://bit.ly/3qqcotZ
๐ Python Online Training: http://bit.ly/3i7DiDD
๐ต AWS Online Training: http://bit.ly/39EOIv9
๐ RPA Online Training: http://bit.ly/39t6MrE
๐ต Data Science Online Training: http://bit.ly/3bywuOu
๐ Big Data Online Training: http://bit.ly/35ADG8L
๐ต Java Online Training: http://bit.ly/35Avwxb
๐ Selenium Online Training: http://bit.ly/2Ltn4ZE
๐ต PMP Online Training: http://bit.ly/2LMP3mT
๐ Tableau Online Training: http://bit.ly/38z4oR9
๐ต Microsoft Azure Online Training: https://bit.ly/3wsAhV4
๐ Power BI Online Training: https://bit.ly/3zq1WHX
---------๐๐๐ฎ๐ซ๐๐ค๐ ๐๐๐ฌ๐ญ๐๐ซ๐ฌ ๐๐ซ๐จ๐ ๐ซ๐๐ฆ๐ฌ---------
๐ต DevOps Engineer Masters Program: http://bit.ly/2MY1OMc
๐ Cloud Architect Masters Program: http://bit.ly/2MY1OMc
๐ต Data Scientist Masters Program: http://bit.ly/3qaVZJA
๐ Big Data Architect Masters Program: http://bit.ly/3qaVZJA
๐ต Machine Learning Engineer Masters Program: http://bit.ly/3bxAhLG
๐ Business Intelligence Masters Program: http://bit.ly/3byBhiS
๐ต Python Developer Masters Program: http://bit.ly/3nw4Rb2
๐ RPA Developer Masters Program: http://bit.ly/3nw4Rb2
-----------------๐๐๐ฎ๐ซ๐๐ค๐ ๐๐๐ ๐๐จ๐ฎ๐ซ๐ฌ๐๐ฌ---------------
๐ตArtificial and Machine Learning PGP: https://bit.ly/2Ziy7b1โ
๐ฃCyberSecurity PGP: https://bit.ly/3eHvI0hโ
๐ตDigital Marketing PGP: https://bit.ly/38cqdnzโ
๐ฃBig Data Engineering PGP: https://bit.ly/3eTSyBCโ
๐ตData Science PGP: https://bit.ly/3dIeYV9โ
๐ฃCloud Computing PGP: https://bit.ly/2B9tHLP
-------------------------------------------------------
About the Python Certification Training by Edureka
This Python course is live, instructor-led & helps you master various Python libraries such as Pandas, Numpy and Matplotlib to name a few, with industry use cases. Enroll now to learn Python online & be a certified Python professional with Edureka.
- - - - - - - - - - - - - - - - - - -
What will you learn in Edurekaโs Python Training
Edurekaโs Python Certification Training will help you to learn more about how to write code in python with examples. Along with that you will also learn more about python syntax, Python basics, python ide, and many more. Not just this, youโll also learn about various other Python Fundamentals in this Python Online Training like:
Introduction to Python
Python Sequences and File Operations
Python Functions
Python OOP
Python Modules
Exception Handling in Python
Python Libraries like NumPy, Pandas, Matplotlib
Python GUI Programming
Computer Vision using Python OpenCV
---------------------------------------------------------------
Who should go for Python Training?
Edurekaโs Python certification course is a good fit for the professionals like Programmers, Developers, Technical Leads, and Architects. Even the developers who are aspiring to be a โMachine Learning Engineer' or the Analytics Managers who are leading a team of analysts can learn Python Programming.
---------------------------------------------------------------
For more information, Please write back to us at sales@edureka.co or call us at IND: 9606058406 / US: 18338555775 (toll-free)
detail
{'title': 'Image Processing Tutorial Using Python | Python OpenCV Tutorial | Python Training | Edureka', 'heatmap': [{'end': 220.91, 'start': 189.53, 'weight': 0.753}, {'end': 720.632, 'start': 660.654, 'weight': 1}], 'summary': "This tutorial covers fundamental image processing concepts, python's role in computer vision, model training using mnist data with keras, cnn creation, deep learning model training with adam optimizer, and image prediction and processing, emphasizing the importance of google colab for model training.", 'chapters': [{'end': 124.69, 'segs': [{'end': 33.335, 'src': 'embed', 'start': 7.349, 'weight': 2, 'content': [{'end': 14.064, 'text': "Hi all, this is Junaid here from Edureka and I welcome you all to this session where we'll be talking about image processing using Python.", 'start': 7.349, 'duration': 6.715}, {'end': 18.128, 'text': "So without any further ado, let me quickly walk you through today's agenda.", 'start': 14.727, 'duration': 3.401}, {'end': 23.551, 'text': 'We start this session by understanding what exactly is image processing and why do we need it?', 'start': 18.689, 'duration': 4.862}, {'end': 24.091, 'text': 'moving ahead?', 'start': 23.551, 'duration': 0.54}, {'end': 27.152, 'text': "I'll be speaking about how python can be used for computer vision.", 'start': 24.191, 'duration': 2.961}, {'end': 33.335, 'text': "Finally, we'll end the session by walking through some of the fundamental concepts that revolve around computer vision,", 'start': 27.872, 'duration': 5.463}], 'summary': 'Junaid from edureka introduces image processing using python and covers its fundamental concepts.', 'duration': 25.986, 'max_score': 7.349, 'thumbnail': 'https://coursnap.oss-ap-southeast-1.aliyuncs.com/video-capture/sfheWK72L74/pics/sfheWK72L747349.jpg'}, {'end': 124.69, 'src': 'embed', 'start': 81.187, 'weight': 0, 'content': [{'end': 84.908, 'text': 'It forms a core research area within the engineering and computer science discipline.', 'start': 81.187, 'duration': 3.721}, {'end': 88.989, 'text': 'You see image processing is used in various fields as you can see here.', 'start': 85.428, 'duration': 3.561}, {'end': 97.452, 'text': 'We can use image processing to track human behavior and their positions this can further be used to develop games or any kind of security task.', 'start': 89.029, 'duration': 8.423}, {'end': 101.644, 'text': 'The another popular use case of image processing is driverless cars.', 'start': 98.223, 'duration': 3.421}, {'end': 104.045, 'text': 'We all are the witness to this.', 'start': 102.124, 'duration': 1.921}, {'end': 108.726, 'text': 'you see, driverless cars are one of the most coolest piece of technology that exists today,', 'start': 104.045, 'duration': 4.681}, {'end': 114.547, 'text': 'and the top players when it comes to driverless cars are like Tesla, Audi, Mercedes and many more.', 'start': 108.726, 'duration': 5.821}, {'end': 120.189, 'text': 'moving on to the last but not least application that is, implementing image processing in medical images,', 'start': 114.547, 'duration': 5.642}, {'end': 124.69, 'text': 'You see we can use this x-ray CT scan or MRI to find abnormalities.', 'start': 120.829, 'duration': 3.861}], 'summary': 'Image processing has diverse applications - tracking human behavior, driverless cars, and medical imaging. tesla, audi, and mercedes are key players in driverless cars.', 'duration': 43.503, 'max_score': 81.187, 'thumbnail': 'https://coursnap.oss-ap-southeast-1.aliyuncs.com/video-capture/sfheWK72L74/pics/sfheWK72L7481187.jpg'}], 'start': 7.349, 'title': 'Image processing with python', 'summary': 'Covers the fundamentals of image processing, the significance of python in computer vision, and its applications in technology, security, and medical imaging.', 'chapters': [{'end': 124.69, 'start': 7.349, 'title': 'Image processing with python', 'summary': 'Discusses the fundamentals of image processing, the significance of python in computer vision, and practical applications in various fields, including technology, security, and medical imaging.', 'duration': 117.341, 'highlights': ['Image processing is a rapidly growing technology used in various fields such as tracking human behavior, developing driverless cars, and medical imaging.', 'Driverless cars are a notable application of image processing, with top players being Tesla, Audi, and Mercedes.', "Python's role in computer vision is discussed, highlighting its relevance and importance in the field."]}], 'duration': 117.341, 'thumbnail': 'https://coursnap.oss-ap-southeast-1.aliyuncs.com/video-capture/sfheWK72L74/pics/sfheWK72L747349.jpg', 'highlights': ['Driverless cars are a notable application of image processing, with top players being Tesla, Audi, and Mercedes.', 'Image processing is a rapidly growing technology used in various fields such as tracking human behavior, developing driverless cars, and medical imaging.', "Python's role in computer vision is discussed, highlighting its relevance and importance in the field."]}, {'end': 919.38, 'segs': [{'end': 151.099, 'src': 'embed', 'start': 125.441, 'weight': 0, 'content': [{'end': 131.546, 'text': 'You see, if a person has a kind of a fracture or a cancer that is being detected in using an MRI,', 'start': 125.441, 'duration': 6.105}, {'end': 138.731, 'text': 'we can actually compare that particular image with respect to the other set of image which does not have or which are totally normal, and then,', 'start': 131.546, 'duration': 7.185}, {'end': 145.717, 'text': 'based on certain criterias and by applying that to a deep learning model, we can figure out whether there is any abnormalities that are present there.', 'start': 138.731, 'duration': 6.986}, {'end': 151.099, 'text': 'But you see we cannot use the output or whatever the opinion given out by a machine learning algorithm.', 'start': 146.317, 'duration': 4.782}], 'summary': 'Comparing mri images with deep learning to detect abnormalities, but not relying solely on machine learning output.', 'duration': 25.658, 'max_score': 125.441, 'thumbnail': 'https://coursnap.oss-ap-southeast-1.aliyuncs.com/video-capture/sfheWK72L74/pics/sfheWK72L74125441.jpg'}, {'end': 220.91, 'src': 'heatmap', 'start': 181.288, 'weight': 1, 'content': [{'end': 188.85, 'text': 'augmentation or whatever it is but we can also process the images using deep learning or machine learning algorithms to perform classification operations.', 'start': 181.288, 'duration': 7.562}, {'end': 194.211, 'text': 'Some of the popular frameworks or libraries here are opencv pytorch.', 'start': 189.53, 'duration': 4.681}, {'end': 197.292, 'text': 'Then we have tensorflow numpy and many more.', 'start': 194.611, 'duration': 2.681}, {'end': 201.395, 'text': 'All right now that we have a brief intuition about image processing.', 'start': 198.473, 'duration': 2.922}, {'end': 209.361, 'text': "Let's see some of the fundamental concepts that revolves around this one of the most fundamental concept in image processing is image filters.", 'start': 201.695, 'duration': 7.666}, {'end': 216.706, 'text': 'You see in image processing filters are mainly used to suppress either the high frequencies in the image, that is, by smoothing an image,', 'start': 209.942, 'duration': 6.764}, {'end': 220.91, 'text': 'or by lowering the frequency, that is, by enhancing or detecting the edges in that particular image.', 'start': 216.706, 'duration': 4.204}], 'summary': 'Image processing involves using deep learning or machine learning algorithms, with popular frameworks like opencv, pytorch, tensorflow, and numpy for tasks such as classification and applying image filters to suppress high frequencies or enhance edges.', 'duration': 39.622, 'max_score': 181.288, 'thumbnail': 'https://coursnap.oss-ap-southeast-1.aliyuncs.com/video-capture/sfheWK72L74/pics/sfheWK72L74181288.jpg'}, {'end': 359.212, 'src': 'embed', 'start': 332.824, 'weight': 4, 'content': [{'end': 339.787, 'text': 'with this, what happens is you know this all of these external layers will be included in our final transformation.', 'start': 332.824, 'duration': 6.963}, {'end': 340.707, 'text': 'All right.', 'start': 340.427, 'duration': 0.28}, {'end': 348.11, 'text': 'So this is exactly why we use padding basically we use padding in order to remove or in order to reduce the loss of information from our image.', 'start': 340.987, 'duration': 7.123}, {'end': 354.208, 'text': 'Moving ahead to the next concept that is thresholding you see thresholding is a type of image segmentation.', 'start': 349.124, 'duration': 5.084}, {'end': 359.212, 'text': 'What happens over here is like when I say image segmentation, as the name suggests here.', 'start': 354.828, 'duration': 4.384}], 'summary': 'Using padding to preserve information in image transformation and thresholding for image segmentation.', 'duration': 26.388, 'max_score': 332.824, 'thumbnail': 'https://coursnap.oss-ap-southeast-1.aliyuncs.com/video-capture/sfheWK72L74/pics/sfheWK72L74332824.jpg'}, {'end': 442.194, 'src': 'embed', 'start': 412.015, 'weight': 5, 'content': [{'end': 415.397, 'text': "We have a man who's taking a like he's trying to record something over here.", 'start': 412.015, 'duration': 3.382}, {'end': 420.78, 'text': 'Now once we are done performing thresholding you can see the image has been converted to now black and white.', 'start': 415.857, 'duration': 4.923}, {'end': 427.305, 'text': 'So with this what happens is it becomes easier to extract noise and also to scan for any abnormalities.', 'start': 421.161, 'duration': 6.144}, {'end': 431.268, 'text': 'Moving ahead to the next topic, that is, connected components.', 'start': 428.206, 'duration': 3.062}, {'end': 434.69, 'text': 'know that we know what does this image thresholding does right?', 'start': 431.268, 'duration': 3.422}, {'end': 438.632, 'text': "So it's basically I know, like either the value can be 1 or 0..", 'start': 434.71, 'duration': 3.922}, {'end': 442.194, 'text': 'So now what happens is we have to process our image right?', 'start': 438.632, 'duration': 3.562}], 'summary': 'Image processing involves thresholding for noise extraction and scanning for abnormalities, followed by processing the image for connected components.', 'duration': 30.179, 'max_score': 412.015, 'thumbnail': 'https://coursnap.oss-ap-southeast-1.aliyuncs.com/video-capture/sfheWK72L74/pics/sfheWK72L74412015.jpg'}, {'end': 472.156, 'src': 'embed', 'start': 446.856, 'weight': 6, 'content': [{'end': 452.38, 'text': 'So here whatever the values were there has been converted to either ones or zeros, but now this is still of no use for us.', 'start': 446.856, 'duration': 5.524}, {'end': 454.702, 'text': 'So this is where the connected components come in.', 'start': 452.74, 'duration': 1.962}, {'end': 462.007, 'text': "So whenever there is the continuous structure of pixels, right? Like as you can see at this guy's hand, this is one continuous row of pixels.", 'start': 455.022, 'duration': 6.985}, {'end': 466.97, 'text': 'Then we have, you can see the other hand, but there is a small white pixel area which is present here, right?', 'start': 462.327, 'duration': 4.643}, {'end': 472.156, 'text': "So if there is a small bridge or there's a small gap here, that means those two are two different components.", 'start': 467.27, 'duration': 4.886}], 'summary': 'Image values converted to ones and zeros, analyzing connected components in images.', 'duration': 25.3, 'max_score': 446.856, 'thumbnail': 'https://coursnap.oss-ap-southeast-1.aliyuncs.com/video-capture/sfheWK72L74/pics/sfheWK72L74446856.jpg'}, {'end': 607.776, 'src': 'embed', 'start': 582.162, 'weight': 2, 'content': [{'end': 586.845, 'text': "Okay, this model will have multiple images and we'll train this model on MNIST data.", 'start': 582.162, 'duration': 4.683}, {'end': 594.047, 'text': 'So in this data is basically a set of handwritten digits and this model.', 'start': 589.724, 'duration': 4.323}, {'end': 600.531, 'text': 'this would be basically on deep learning model that we are going to create on our own, and the way this works is, once we train this model,', 'start': 594.047, 'duration': 6.484}, {'end': 604.074, 'text': 'whatever input image I provide, which the size would be 28 comma 28..', 'start': 600.531, 'duration': 3.543}, {'end': 607.776, 'text': 'It will give me a correct prediction what that image exactly is doing.', 'start': 604.074, 'duration': 3.702}], 'summary': 'Creating a deep learning model to predict handwritten digits with mnist data.', 'duration': 25.614, 'max_score': 582.162, 'thumbnail': 'https://coursnap.oss-ap-southeast-1.aliyuncs.com/video-capture/sfheWK72L74/pics/sfheWK72L74582162.jpg'}, {'end': 650.718, 'src': 'embed', 'start': 626.383, 'weight': 8, 'content': [{'end': 632.147, 'text': 'Okay, and inside this board anytime, like you, take up any handwritten image and you press 9,', 'start': 626.383, 'duration': 5.764}, {'end': 639.111, 'text': 'what happens is we will get a bounding box which will detect this particular number, that is 9, and give me an output over here.', 'start': 632.147, 'duration': 6.964}, {'end': 646.435, 'text': 'All right, and if I write 8 it will just detect this and give me a number 8 and all of this will be happening in real time.', 'start': 639.711, 'duration': 6.724}, {'end': 648.416, 'text': 'Okay So this sounds pretty interesting.', 'start': 646.775, 'duration': 1.641}, {'end': 650.718, 'text': 'So let us now quickly start working on this.', 'start': 648.476, 'duration': 2.242}], 'summary': 'Real-time handwritten number detection with bounding boxes for 8 and 9.', 'duration': 24.335, 'max_score': 626.383, 'thumbnail': 'https://coursnap.oss-ap-southeast-1.aliyuncs.com/video-capture/sfheWK72L74/pics/sfheWK72L74626383.jpg'}, {'end': 720.632, 'src': 'heatmap', 'start': 660.654, 'weight': 1, 'content': [{'end': 662.835, 'text': 'This images is nothing but MNIST image.', 'start': 660.654, 'duration': 2.181}, {'end': 664.736, 'text': 'So as you can see, this is a small image.', 'start': 663.115, 'duration': 1.621}, {'end': 669.697, 'text': "It's a seven and then if I put like five, okay, so this image represents three.", 'start': 665.076, 'duration': 4.621}, {'end': 675.038, 'text': 'So this is basically a small data set which I have done and now inside this will train a model.', 'start': 669.737, 'duration': 5.301}, {'end': 677.419, 'text': "So let's quickly get that done.", 'start': 675.318, 'duration': 2.101}, {'end': 678.999, 'text': 'So to open a Jupiter notebook here.', 'start': 677.479, 'duration': 1.52}, {'end': 687.65, 'text': 'We just need to press ctrl alt P All right, and now we just hit that and now if I will save this.', 'start': 679.379, 'duration': 8.271}, {'end': 690.812, 'text': 'All right.', 'start': 690.552, 'duration': 0.26}, {'end': 697.018, 'text': "So now we are supposed to train this particular model, right? So what I'm going to do here is we'll first try saving this up.", 'start': 690.872, 'duration': 6.146}, {'end': 704.043, 'text': "So I'll save this inside this model development and I'll just give a name here as model.", 'start': 698.259, 'duration': 5.784}, {'end': 707.706, 'text': "Development And we'll save this.", 'start': 705.344, 'duration': 2.362}, {'end': 708.407, 'text': 'All right.', 'start': 708.167, 'duration': 0.24}, {'end': 711.345, 'text': 'All right guys.', 'start': 710.905, 'duration': 0.44}, {'end': 715.748, 'text': 'So, now that we are here at our Jupiter notebook first off, we have to install a couple of dependencies right?', 'start': 711.385, 'duration': 4.363}, {'end': 720.632, 'text': "So in order to install, I don't have to actually go back to my anaconda command prompt and type.", 'start': 716.029, 'duration': 4.603}], 'summary': 'The transcript discusses working with a small mnist image dataset and training a model in a jupyter notebook, involving saving and installing dependencies.', 'duration': 59.978, 'max_score': 660.654, 'thumbnail': 'https://coursnap.oss-ap-southeast-1.aliyuncs.com/video-capture/sfheWK72L74/pics/sfheWK72L74660654.jpg'}, {'end': 771.467, 'src': 'embed', 'start': 744.946, 'weight': 7, 'content': [{'end': 749.01, 'text': 'So as you can see here, we have successfully installed Keras in our jupyter notebook.', 'start': 744.946, 'duration': 4.064}, {'end': 751.972, 'text': 'This would usually take time depending upon your internet speed.', 'start': 749.53, 'duration': 2.442}, {'end': 755.314, 'text': 'All right, so let me quickly import Keras now.', 'start': 752.272, 'duration': 3.042}, {'end': 771.467, 'text': 'So we have from Keras dot data set import MNIST data, right? And then a couple more things that we I would be needing is like numpy import numpy.', 'start': 756.856, 'duration': 14.611}], 'summary': 'Successfully installed keras in jupyter notebook, featuring mnist dataset import and numpy.', 'duration': 26.521, 'max_score': 744.946, 'thumbnail': 'https://coursnap.oss-ap-southeast-1.aliyuncs.com/video-capture/sfheWK72L74/pics/sfheWK72L74744946.jpg'}], 'start': 125.441, 'title': 'Image processing and model training', 'summary': 'Discusses fundamentals of image processing in medical diagnosis and explains concepts such as image filters, padding, thresholding, and connected components. it also covers the implementation of image processing and model training using mnist data and keras in jupyter notebook, including the creation of a deep learning model and the installation of required dependencies.', 'chapters': [{'end': 517.26, 'start': 125.441, 'title': 'Fundamentals of image processing', 'summary': 'Discusses the application of image processing in medical diagnosis and explains fundamental concepts such as image filters, padding, thresholding, and connected components.', 'duration': 391.819, 'highlights': ['Image processing is used in medical diagnosis to compare MRI images and detect abnormalities using deep learning models.', 'Python is a popular language for image processing, with libraries like OpenCV, PyTorch, TensorFlow, and NumPy for performing classification operations.', 'Image filters are fundamental in image processing, used to suppress high frequencies, smooth images, enhance edges, and can be applied in frequency or spatial domains.', 'Padding is used in image processing to prevent loss of information by adding extra layer of pixels to the image.', 'Thresholding involves setting a threshold value to convert an image to black and white, making it easier to extract noise and scan for abnormalities.', 'Connected components are used to process thresholded images by identifying continuous structures of pixels and performing analysis and operations such as building bounding boxes.']}, {'end': 919.38, 'start': 517.26, 'title': 'Image processing and model training', 'summary': 'Covers the implementation of image processing and model training using mnist data and keras in jupyter notebook, including the creation of a deep learning model and the installation of required dependencies.', 'duration': 402.12, 'highlights': ['The model created will be trained on MNIST data, a set of handwritten digits, using a deep learning approach, and will provide correct predictions for input images of size 28x28.', 'The implementation involves creating a board where handwritten images can be input, and real-time detection of the handwritten number will be displayed, such as detecting the number 9 and providing an output for it.', 'Dependencies like Keras and scikit-learn are installed using Jupyter notebook, with a demonstration of successful installation of Keras in the notebook.', 'The process involves loading MNIST data, displaying the images, and their corresponding labels using matplotlib in the Jupyter notebook to visualize the dataset.', 'The chapter also includes the detection of individual handwritten numbers and the prevention of color conversion while displaying the images using matplotlib.']}], 'duration': 793.939, 'thumbnail': 'https://coursnap.oss-ap-southeast-1.aliyuncs.com/video-capture/sfheWK72L74/pics/sfheWK72L74125441.jpg', 'highlights': ['Image processing is used in medical diagnosis to compare MRI images and detect abnormalities using deep learning models.', 'Python is a popular language for image processing, with libraries like OpenCV, PyTorch, TensorFlow, and NumPy for performing classification operations.', 'The model created will be trained on MNIST data, a set of handwritten digits, using a deep learning approach, and will provide correct predictions for input images of size 28x28.', 'Image filters are fundamental in image processing, used to suppress high frequencies, smooth images, enhance edges, and can be applied in frequency or spatial domains.', 'Padding is used in image processing to prevent loss of information by adding an extra layer of pixels to the image.', 'Thresholding involves setting a threshold value to convert an image to black and white, making it easier to extract noise and scan for abnormalities.', 'Connected components are used to process thresholded images by identifying continuous structures of pixels and performing analysis and operations such as building bounding boxes.', 'Dependencies like Keras and scikit-learn are installed using Jupyter notebook, with a demonstration of successful installation of Keras in the notebook.', 'The implementation involves creating a board where handwritten images can be input, and real-time detection of the handwritten number will be displayed, such as detecting the number 9 and providing an output for it.', 'The process involves loading MNIST data, displaying the images, and their corresponding labels using matplotlib in the Jupyter notebook to visualize the dataset.']}, {'end': 1217.391, 'segs': [{'end': 966.528, 'src': 'embed', 'start': 920.941, 'weight': 0, 'content': [{'end': 922.622, 'text': 'So now our image will be in grayscale.', 'start': 920.941, 'duration': 1.681}, {'end': 923.343, 'text': 'All right.', 'start': 923.063, 'duration': 0.28}, {'end': 926.446, 'text': "So now that we know how our data is let's now create a model.", 'start': 923.723, 'duration': 2.723}, {'end': 931.666, 'text': "So what I'm going to do is I'll just give a comment here creating model.", 'start': 927.443, 'duration': 4.223}, {'end': 939.872, 'text': "Okay So in order to create model if you know a bit of how deep learning works here, we'll be using something called as Keras framework.", 'start': 933.247, 'duration': 6.625}, {'end': 942.374, 'text': 'All right and the way this works is.', 'start': 940.392, 'duration': 1.982}, {'end': 948.678, 'text': "we'll be having multiple layers and each of these layers will have, you know, a convolution layer to take up extra features.", 'start': 942.374, 'duration': 6.304}, {'end': 952.1, 'text': 'If you have to reduce the dimensions will be then using filters for this.', 'start': 949.058, 'duration': 3.042}, {'end': 957.004, 'text': "So let's now see how we can do that from Keras dot layers.", 'start': 952.501, 'duration': 4.503}, {'end': 965.127, 'text': 'import dense layer all right, and then we have convolution layer.', 'start': 958.726, 'duration': 6.401}, {'end': 966.528, 'text': 'so this is basically a filter.', 'start': 965.127, 'duration': 1.401}], 'summary': 'Creating a model using keras framework with multiple layers and convolution filters.', 'duration': 45.587, 'max_score': 920.941, 'thumbnail': 'https://coursnap.oss-ap-southeast-1.aliyuncs.com/video-capture/sfheWK72L74/pics/sfheWK72L74920941.jpg'}, {'end': 1127.514, 'src': 'embed', 'start': 1082.023, 'weight': 1, 'content': [{'end': 1087.206, 'text': "Okay So as this is just a convolution 2D after this we'll have something called as average pool.", 'start': 1082.023, 'duration': 5.183}, {'end': 1090.407, 'text': 'This is done usually to decrease your dimensions.', 'start': 1087.226, 'duration': 3.181}, {'end': 1096.73, 'text': "Okay, so model dot add so we'll add average pool and now we'll just have to provide a pool size.", 'start': 1090.507, 'duration': 6.223}, {'end': 1106.889, 'text': "So we'll be just put here pool size and this is usually 3 cross 3 as or 5 cross 5 but as is a small image will just give 2 cross 2.", 'start': 1097.947, 'duration': 8.942}, {'end': 1107.109, 'text': 'All right.', 'start': 1106.889, 'duration': 0.22}, {'end': 1107.929, 'text': 'So this is done.', 'start': 1107.149, 'duration': 0.78}, {'end': 1112.31, 'text': 'We are done with one particular layer now after this will again pass our convolution layer.', 'start': 1107.989, 'duration': 4.321}, {'end': 1117.992, 'text': "So I'll just copy this over here and paste it and I'll just do couple changes instead of input shape.", 'start': 1112.33, 'duration': 5.662}, {'end': 1120.872, 'text': "I'll just replace this by nothing.", 'start': 1118.352, 'duration': 2.52}, {'end': 1122.833, 'text': 'and then the kernel size we can just give.', 'start': 1120.872, 'duration': 1.961}, {'end': 1127.514, 'text': 'instead of 5, cross 5, we can give 7, cross 7 or 8 cross 8 totally depends upon you.', 'start': 1122.833, 'duration': 4.681}], 'summary': 'A 2d convolution layer is added, followed by an average pooling layer with a 2x2 pool size. another convolution layer is then added with variable kernel sizes.', 'duration': 45.491, 'max_score': 1082.023, 'thumbnail': 'https://coursnap.oss-ap-southeast-1.aliyuncs.com/video-capture/sfheWK72L74/pics/sfheWK72L741082023.jpg'}, {'end': 1181.825, 'src': 'embed', 'start': 1151.799, 'weight': 3, 'content': [{'end': 1153.16, 'text': "So here we'll have units.", 'start': 1151.799, 'duration': 1.361}, {'end': 1159.085, 'text': 'So as this performing a classification task, so units here represents how many number of outputs we have,', 'start': 1153.4, 'duration': 5.685}, {'end': 1162.389, 'text': 'as we know that the number of outputs here range from 0 to 9..', 'start': 1159.085, 'duration': 3.304}, {'end': 1167.834, 'text': "Okay, because that's the numbers we have right? So we'll give here units as 10.", 'start': 1162.389, 'duration': 5.445}, {'end': 1170.196, 'text': 'All right, and then we have activation.', 'start': 1167.834, 'duration': 2.362}, {'end': 1173.139, 'text': 'So the activation that we are going to use is softmax.', 'start': 1171.017, 'duration': 2.122}, {'end': 1178.423, 'text': 'right. so let us now compile this or before that.', 'start': 1174.741, 'duration': 3.682}, {'end': 1181.825, 'text': "let me run this cell and finally we'll compile it.", 'start': 1178.423, 'duration': 3.402}], 'summary': 'Performing a classification task with 10 output units using softmax activation.', 'duration': 30.026, 'max_score': 1151.799, 'thumbnail': 'https://coursnap.oss-ap-southeast-1.aliyuncs.com/video-capture/sfheWK72L74/pics/sfheWK72L741151799.jpg'}, {'end': 1224.594, 'src': 'embed', 'start': 1196.849, 'weight': 4, 'content': [{'end': 1201.192, 'text': "If I use past categorical cross entropy, then I don't have to convert my outputs into one hot layers.", 'start': 1196.849, 'duration': 4.343}, {'end': 1203.373, 'text': "Okay, so I'm going to do that.", 'start': 1201.772, 'duration': 1.601}, {'end': 1207.075, 'text': 'So it will be sparse categorical cross entropy.', 'start': 1203.433, 'duration': 3.642}, {'end': 1212.198, 'text': "All right, and the optimizer that I'm going to use here is Adam.", 'start': 1209.276, 'duration': 2.922}, {'end': 1215.09, 'text': 'You can use anything that you want.', 'start': 1213.749, 'duration': 1.341}, {'end': 1217.391, 'text': 'You can use like stochastic gradient descent.', 'start': 1215.13, 'duration': 2.261}, {'end': 1224.594, 'text': 'Then you have various types that you can use the best one that is that is as of now is Adam and then metrics that we want to represent.', 'start': 1217.731, 'duration': 6.863}], 'summary': 'Using sparse categorical cross entropy and adam optimizer for metrics.', 'duration': 27.745, 'max_score': 1196.849, 'thumbnail': 'https://coursnap.oss-ap-southeast-1.aliyuncs.com/video-capture/sfheWK72L74/pics/sfheWK72L741196849.jpg'}], 'start': 920.941, 'title': 'Convolutional neural network model creation', 'summary': 'Discusses creating a cnn model using keras framework with convolution, filter size, input shape, and activation function, and covers the addition of convolution and pooling layers with specific configurations, dense layer addition, and loss and optimizer usage.', 'chapters': [{'end': 1081.923, 'start': 920.941, 'title': 'Creating a convolutional neural network model', 'summary': 'Discusses the process of creating a convolutional neural network model using keras framework, involving the use of convolution, filter size, input shape, and activation function to enhance non-linearity.', 'duration': 160.982, 'highlights': ['The process of creating a Convolutional Neural Network model using Keras framework is discussed, involving the use of convolution, filter size, input shape, and activation function to enhance non-linearity.', 'The use of Keras framework and sequential model for creating the CNN model is explained.', 'Explanation of the use of convolution layer, filter size, and input shape for the CNN model.']}, {'end': 1217.391, 'start': 1082.023, 'title': 'Convolution and pooling layers in cnn', 'summary': 'Covers the addition of convolution and average pooling layers with pool size 2x2, followed by another convolution layer with a kernel size of 7x7, flattening the model, and adding a dense layer with 10 units and softmax activation, using the sparse categorical cross entropy loss and adam optimizer.', 'duration': 135.368, 'highlights': ['Adding convolution and average pooling layers with pool size 2x2', 'Adding another convolution layer with a kernel size of 7x7', 'Flattening the model and adding a dense layer with 10 units and softmax activation', 'Using sparse categorical cross entropy loss and Adam optimizer']}], 'duration': 296.45, 'thumbnail': 'https://coursnap.oss-ap-southeast-1.aliyuncs.com/video-capture/sfheWK72L74/pics/sfheWK72L74920941.jpg', 'highlights': ['The process of creating a Convolutional Neural Network model using Keras framework is discussed, involving the use of convolution, filter size, input shape, and activation function to enhance non-linearity.', 'Adding convolution and average pooling layers with pool size 2x2', 'Adding another convolution layer with a kernel size of 7x7', 'Flattening the model and adding a dense layer with 10 units and softmax activation', 'Using sparse categorical cross entropy loss and Adam optimizer', 'The use of Keras framework and sequential model for creating the CNN model is explained.', 'Explanation of the use of convolution layer, filter size, and input shape for the CNN model.']}, {'end': 1468.689, 'segs': [{'end': 1263.48, 'src': 'embed', 'start': 1217.731, 'weight': 0, 'content': [{'end': 1224.594, 'text': 'Then you have various types that you can use the best one that is that is as of now is Adam and then metrics that we want to represent.', 'start': 1217.731, 'duration': 6.863}, {'end': 1226.575, 'text': 'This is by accuracy.', 'start': 1224.614, 'duration': 1.961}, {'end': 1232.158, 'text': 'All right, so this is done and now finally we will compile our model.', 'start': 1227.916, 'duration': 4.242}, {'end': 1233.659, 'text': 'So let me execute this.', 'start': 1232.678, 'duration': 0.981}, {'end': 1238.061, 'text': 'All right, maybe we have done a small typo here.', 'start': 1235.959, 'duration': 2.102}, {'end': 1241.523, 'text': 'So let me change this from accuracy to ACC to accuracy.', 'start': 1238.601, 'duration': 2.922}, {'end': 1244.726, 'text': "Oh, yeah, it's going to be optimized right so optimize.", 'start': 1241.543, 'duration': 3.183}, {'end': 1245.967, 'text': 'All right.', 'start': 1245.706, 'duration': 0.261}, {'end': 1249.569, 'text': "So now what we're going to do is finally we are going to see the model dot summary.", 'start': 1246.267, 'duration': 3.302}, {'end': 1251.05, 'text': 'So model dot summary.', 'start': 1249.709, 'duration': 1.341}, {'end': 1254.413, 'text': 'This just basically gives us what we have included here.', 'start': 1251.711, 'duration': 2.702}, {'end': 1260.538, 'text': 'So as you can see we have total number of layers as 1446 the number of parameters.', 'start': 1255.013, 'duration': 5.525}, {'end': 1263.48, 'text': "It's pretty small when you compare to our deep learning applications.", 'start': 1260.598, 'duration': 2.882}], 'summary': 'Utilizing adam optimizer with accuracy metrics, compiling a model with 1446 layers and a small number of parameters.', 'duration': 45.749, 'max_score': 1217.731, 'thumbnail': 'https://coursnap.oss-ap-southeast-1.aliyuncs.com/video-capture/sfheWK72L74/pics/sfheWK72L741217731.jpg'}, {'end': 1305.828, 'src': 'embed', 'start': 1282.139, 'weight': 2, 'content': [{'end': 1289.905, 'text': "It's q1 because if we give more it's going to over fist right? So number of epochs, this would be one and the batch size will just give one.", 'start': 1282.139, 'duration': 7.766}, {'end': 1295.5, 'text': 'So in one go it will just take up all the values right and let me execute this.', 'start': 1291.957, 'duration': 3.543}, {'end': 1300.744, 'text': 'All right, so we basically have to reshape our image.', 'start': 1297.982, 'duration': 2.762}, {'end': 1302.765, 'text': 'So let me quickly get that done as well.', 'start': 1301.124, 'duration': 1.641}, {'end': 1305.828, 'text': 'Okay So basically we have to pre-process our image.', 'start': 1302.885, 'duration': 2.943}], 'summary': 'Adjusting number of epochs and batch size for image reshaping and preprocessing.', 'duration': 23.689, 'max_score': 1282.139, 'thumbnail': 'https://coursnap.oss-ap-southeast-1.aliyuncs.com/video-capture/sfheWK72L74/pics/sfheWK72L741282139.jpg'}, {'end': 1358.733, 'src': 'embed', 'start': 1327.753, 'weight': 1, 'content': [{'end': 1335.375, 'text': 'So what this would do is like we will just get down here add one more column or so.', 'start': 1327.753, 'duration': 7.622}, {'end': 1342.856, 'text': "Okay, and now we'll take X train should be equal to X train dot reshape.", 'start': 1336.335, 'duration': 6.521}, {'end': 1348.338, 'text': "All right, and now we'll pass a tuple value which gives us this thing.", 'start': 1343.857, 'duration': 4.481}, {'end': 1350.35, 'text': 'Okay, so minus 1.', 'start': 1348.518, 'duration': 1.832}, {'end': 1354.672, 'text': 'comma 28 comma 28 comma 1.', 'start': 1350.35, 'duration': 4.322}, {'end': 1357.873, 'text': "So if I see the shape now you'll see here for this.", 'start': 1354.672, 'duration': 3.201}, {'end': 1358.733, 'text': 'Yeah, you have one.', 'start': 1358.053, 'duration': 0.68}], 'summary': 'Reshaping x train to 28x28x1, resulting in shape (1)', 'duration': 30.98, 'max_score': 1327.753, 'thumbnail': 'https://coursnap.oss-ap-southeast-1.aliyuncs.com/video-capture/sfheWK72L74/pics/sfheWK72L741327753.jpg'}, {'end': 1468.689, 'src': 'embed', 'start': 1430.649, 'weight': 5, 'content': [{'end': 1434.311, 'text': "Okay So for this what we have what we'll do is we'll rerun this model from start.", 'start': 1430.649, 'duration': 3.662}, {'end': 1436.732, 'text': "Okay, so we'll run this from here.", 'start': 1434.351, 'duration': 2.381}, {'end': 1438.653, 'text': 'Yeah, yeah, yeah.', 'start': 1437.053, 'duration': 1.6}, {'end': 1445.577, 'text': 'So, as you can see here now, our model is working, and this currently is in our training phase right now,', 'start': 1440.474, 'duration': 5.103}, {'end': 1449.559, 'text': 'and the way this thing trains totally depends upon your CPU or a compute power.', 'start': 1445.577, 'duration': 3.982}, {'end': 1453.542, 'text': 'What people usually do is you know if they have any machine learning or deep learning model?', 'start': 1450.14, 'duration': 3.402}, {'end': 1456.243, 'text': "they would just go down to Google collab, as it's totally free of cost.", 'start': 1453.542, 'duration': 2.701}, {'end': 1459.325, 'text': "Okay So let's now wait for this thing to train off.", 'start': 1456.823, 'duration': 2.502}, {'end': 1468.689, 'text': 'All right, so we have now successfully trained our model.', 'start': 1466.328, 'duration': 2.361}], 'summary': 'Model successfully trained using cpu power, free on google colab.', 'duration': 38.04, 'max_score': 1430.649, 'thumbnail': 'https://coursnap.oss-ap-southeast-1.aliyuncs.com/video-capture/sfheWK72L74/pics/sfheWK72L741430649.jpg'}], 'start': 1217.731, 'title': 'Deep learning model training and reshaping data for image classification', 'summary': 'Covers training a deep learning model using adam optimizer with a small number of parameters, 1 epoch, and batch size of 1, and the need to pre-process the image. it also explains reshaping data for image classification, emphasizing specific dimensions (1, 28, 28, 1), and the importance of using google colab for training machine learning models.', 'chapters': [{'end': 1305.828, 'start': 1217.731, 'title': 'Deep learning model training', 'summary': 'Covers the training of a deep learning model using adam optimizer, achieving a small number of parameters, and training with 1 epoch and batch size of 1, with the need to pre-process the image.', 'duration': 88.097, 'highlights': ['The model consists of 1446 layers with a small number of parameters compared to other deep learning applications, as indicated by the model summary.', 'The model is trained with 1 epoch and a batch size of 1 to avoid overfitting.', 'The chapter emphasizes using the Adam optimizer for training the model.', 'Pre-processing of the image is required before training the model.']}, {'end': 1468.689, 'start': 1306.648, 'title': 'Reshaping data for image classification', 'summary': 'Explains the process of reshaping data for image classification, emphasizing the need for specific dimensions (1, 28, 28, 1), and highlights the importance of using google colab for training machine learning models.', 'duration': 162.041, 'highlights': ['The need to reshape the data to include one more column or dimension, with the specific tuple value (1, 28, 28, 1) for the shape, as it is essential for image classification.', 'The importance of utilizing Google Colab for training machine learning or deep learning models, as it is a cost-effective option and the training process depends on CPU or compute power.', 'The process of rerunning the model fitting phase to ensure the model is working correctly and has been successfully trained.']}], 'duration': 250.958, 'thumbnail': 'https://coursnap.oss-ap-southeast-1.aliyuncs.com/video-capture/sfheWK72L74/pics/sfheWK72L741217731.jpg', 'highlights': ['The model consists of 1446 layers with a small number of parameters compared to other deep learning applications, as indicated by the model summary.', 'The need to reshape the data to include one more column or dimension, with the specific tuple value (1, 28, 28, 1) for the shape, as it is essential for image classification.', 'The model is trained with 1 epoch and a batch size of 1 to avoid overfitting.', 'The chapter emphasizes using the Adam optimizer for training the model.', 'Pre-processing of the image is required before training the model.', 'The importance of utilizing Google Colab for training machine learning or deep learning models, as it is a cost-effective option and the training process depends on CPU or compute power.', 'The process of rerunning the model fitting phase to ensure the model is working correctly and has been successfully trained.']}, {'end': 2053.514, 'segs': [{'end': 1516.484, 'src': 'embed', 'start': 1488.378, 'weight': 2, 'content': [{'end': 1494.37, 'text': "So now what I'm going to do is import CV to All right.", 'start': 1488.378, 'duration': 5.992}, {'end': 1501.294, 'text': 'Now with CV2, I can just provide a path or I can also have something like pandas.', 'start': 1494.63, 'duration': 6.664}, {'end': 1508.038, 'text': "Okay And now what I'm going to do is I want to read this particular image, right? So pandas.", 'start': 1502.875, 'duration': 5.163}, {'end': 1508.959, 'text': 'All right.', 'start': 1508.719, 'duration': 0.24}, {'end': 1510.82, 'text': 'Apart from pandas, I can just use CV2.', 'start': 1509.199, 'duration': 1.621}, {'end': 1512.801, 'text': 'So let me quickly import that.', 'start': 1511.18, 'duration': 1.621}, {'end': 1514.542, 'text': "So I'll have CV2.", 'start': 1513.402, 'duration': 1.14}, {'end': 1516.484, 'text': 'All right.', 'start': 1516.223, 'duration': 0.261}], 'summary': 'Import cv2 to read and process images using pandas and cv2.', 'duration': 28.106, 'max_score': 1488.378, 'thumbnail': 'https://coursnap.oss-ap-southeast-1.aliyuncs.com/video-capture/sfheWK72L74/pics/sfheWK72L741488378.jpg'}, {'end': 1578.035, 'src': 'embed', 'start': 1548.474, 'weight': 3, 'content': [{'end': 1552.797, 'text': "Okay The reason why I'm getting an error is because obviously I'm supposed to give a path like this.", 'start': 1548.474, 'duration': 4.323}, {'end': 1556.018, 'text': 'And yeah, so it should work now.', 'start': 1554.236, 'duration': 1.782}, {'end': 1558.199, 'text': 'All right.', 'start': 1557.939, 'duration': 0.26}, {'end': 1561.061, 'text': "So now we'll try reading this image over here.", 'start': 1558.459, 'duration': 2.602}, {'end': 1567.667, 'text': 'All right.', 'start': 1567.446, 'duration': 0.221}, {'end': 1571.149, 'text': "So the reason why we are getting this error is because it's kind of unable to read this.", 'start': 1567.687, 'duration': 3.462}, {'end': 1578.035, 'text': 'So sometimes what happens is, you know, depending upon the system, we have to use different orientation of this particular path.', 'start': 1571.61, 'duration': 6.425}], 'summary': 'Error occurred due to incorrect path format, requiring different orientation.', 'duration': 29.561, 'max_score': 1548.474, 'thumbnail': 'https://coursnap.oss-ap-southeast-1.aliyuncs.com/video-capture/sfheWK72L74/pics/sfheWK72L741548474.jpg'}, {'end': 1699.821, 'src': 'embed', 'start': 1670.764, 'weight': 0, 'content': [{'end': 1675.005, 'text': 'We have Keras function which says model dot save.', 'start': 1670.764, 'duration': 4.241}, {'end': 1682.69, 'text': 'All right, and now what this model or save do is we have to pass a path and all we need to do is like my model.', 'start': 1676.027, 'duration': 6.663}, {'end': 1689.014, 'text': 'H5 and now once execute this you can see I will have an H5 file.', 'start': 1684.351, 'duration': 4.663}, {'end': 1690.195, 'text': 'All right.', 'start': 1689.915, 'duration': 0.28}, {'end': 1694.238, 'text': "So this is how we work and train a model know what I'm going to do is in our next stage.", 'start': 1690.415, 'duration': 3.823}, {'end': 1699.821, 'text': "We will get back to a python and we'll just pass a value like like this particular image.", 'start': 1694.658, 'duration': 5.163}], 'summary': 'Using keras function model.save to create an h5 file for model training.', 'duration': 29.057, 'max_score': 1670.764, 'thumbnail': 'https://coursnap.oss-ap-southeast-1.aliyuncs.com/video-capture/sfheWK72L74/pics/sfheWK72L741670764.jpg'}, {'end': 1771.266, 'src': 'embed', 'start': 1743.243, 'weight': 1, 'content': [{'end': 1747.327, 'text': "So what we're going to do is we're going to use something called as pygame module and pygame module, if you know,", 'start': 1743.243, 'duration': 4.084}, {'end': 1750.89, 'text': "it's one of the most popular application or a framework to develop games.", 'start': 1747.327, 'duration': 3.563}, {'end': 1756.775, 'text': "Okay, So the reason why I'm going to use pygame is because here we have something called as event loop, which is a constantly running loop,", 'start': 1751.47, 'duration': 5.305}, {'end': 1758.617, 'text': "until unless we don't close our application window.", 'start': 1756.775, 'duration': 1.842}, {'end': 1761.259, 'text': 'So that means our window is continuously getting fresh.', 'start': 1758.997, 'duration': 2.262}, {'end': 1763.962, 'text': 'So let me quickly show you how we can work over here.', 'start': 1761.72, 'duration': 2.242}, {'end': 1771.266, 'text': "So, first off, we'll import pygame and then system framework.", 'start': 1764.322, 'duration': 6.944}], 'summary': 'Using pygame module for game development, leveraging event loop for continuous window refresh.', 'duration': 28.023, 'max_score': 1743.243, 'thumbnail': 'https://coursnap.oss-ap-southeast-1.aliyuncs.com/video-capture/sfheWK72L74/pics/sfheWK72L741743243.jpg'}], 'start': 1469.049, 'title': 'Image prediction and python model training', 'summary': 'Covers image loading and errors in opencv, and training a model using keras for continuous image prediction and drawing with pygame.', 'chapters': [{'end': 1571.149, 'start': 1469.049, 'title': 'Image prediction with opencv', 'summary': 'Covers the process of loading and reading images using opencv, encountering errors while reading the image, and the need to provide the correct path for successful image reading.', 'duration': 102.1, 'highlights': ['The process of loading and reading images using OpenCV is demonstrated, including providing the correct path for successful image reading.', 'Encountering errors while reading the image is highlighted, showcasing the need to provide a correct path for successful image reading.']}, {'end': 2053.514, 'start': 1571.61, 'title': 'Python model training and application development', 'summary': 'Details the process of training and saving a model using keras, then demonstrates the development of a python application with pygame for continuous image prediction and drawing on a display surface.', 'duration': 481.904, 'highlights': ['The chapter demonstrates training a model using Keras and saving it as an H5 file.', 'The chapter showcases the development of a Python application with Pygame for continuous image prediction and drawing on a display surface.']}], 'duration': 584.465, 'thumbnail': 'https://coursnap.oss-ap-southeast-1.aliyuncs.com/video-capture/sfheWK72L74/pics/sfheWK72L741469049.jpg', 'highlights': ['The chapter demonstrates training a model using Keras and saving it as an H5 file.', 'The chapter showcases the development of a Python application with Pygame for continuous image prediction and drawing on a display surface.', 'The process of loading and reading images using OpenCV is demonstrated, including providing the correct path for successful image reading.', 'Encountering errors while reading the image is highlighted, showcasing the need to provide a correct path for successful image reading.']}, {'end': 2748.933, 'segs': [{'end': 2126.041, 'src': 'embed', 'start': 2053.614, 'weight': 2, 'content': [{'end': 2065.217, 'text': 'So it will be x chord and y chord and finally, like shape or the size of that, and 0 and finally will append the coordinates to our.', 'start': 2053.614, 'duration': 11.603}, {'end': 2066.879, 'text': 'so number x chord dot append.', 'start': 2065.217, 'duration': 1.662}, {'end': 2071.695, 'text': "We'll give you a regex cord and similarly for white cord.", 'start': 2068.732, 'duration': 2.963}, {'end': 2074.697, 'text': 'All right.', 'start': 2074.438, 'duration': 0.259}, {'end': 2076.159, 'text': 'So this is one thing.', 'start': 2074.938, 'duration': 1.221}, {'end': 2083.428, 'text': 'So next off if you have if event not type and now is writing will be true.', 'start': 2076.4, 'duration': 7.028}, {'end': 2087.272, 'text': "Okay So here we'll have mouse button.", 'start': 2083.547, 'duration': 3.725}, {'end': 2090.629, 'text': "All right, so it's going to be mouse button up.", 'start': 2088.928, 'duration': 1.701}, {'end': 2094.652, 'text': 'So when the mouse button is up, we have to do perform various tasks.', 'start': 2091.21, 'duration': 3.442}, {'end': 2100.396, 'text': 'That is nothing but once we are done writing, right? We have to whatever the values that we have we have to take it and pass it to a model.', 'start': 2094.692, 'duration': 5.704}, {'end': 2104.539, 'text': "But before that, let's also do it for mouse button up or when the mouse button is down.", 'start': 2100.656, 'duration': 3.883}, {'end': 2115.386, 'text': 'So if even dot type is equal equal to mouse button up or mouse button down that is This is going to be down.', 'start': 2104.659, 'duration': 10.727}, {'end': 2120.158, 'text': 'Okay So when a person starts writing then what happens is is writing variable.', 'start': 2116.656, 'duration': 3.502}, {'end': 2120.858, 'text': 'This would be true.', 'start': 2120.198, 'duration': 0.66}, {'end': 2123.84, 'text': 'And now coming down to mouse button up.', 'start': 2122.139, 'duration': 1.701}, {'end': 2126.041, 'text': 'So this is a pretty important part here.', 'start': 2124.1, 'duration': 1.941}], 'summary': 'Code involves mouse button events and coordinates for model input.', 'duration': 72.427, 'max_score': 2053.614, 'thumbnail': 'https://coursnap.oss-ap-southeast-1.aliyuncs.com/video-capture/sfheWK72L74/pics/sfheWK72L742053614.jpg'}, {'end': 2694.42, 'src': 'embed', 'start': 2669.869, 'weight': 0, 'content': [{'end': 2675.611, 'text': 'So if would 1 right? So it is it is representing 2 because we have this curve as is a deep learning model.', 'start': 2669.869, 'duration': 5.742}, {'end': 2677.191, 'text': "It's more of a pattern.", 'start': 2675.731, 'duration': 1.46}, {'end': 2680.912, 'text': "Okay So now if I have to represent 8, it's something like this.", 'start': 2677.471, 'duration': 3.441}, {'end': 2687.574, 'text': "Okay, so it's showing 2 but if I put it like this 8.", 'start': 2681.333, 'duration': 6.241}, {'end': 2689.896, 'text': "All right, so it's all about patterns.", 'start': 2687.574, 'duration': 2.322}, {'end': 2692.398, 'text': 'Okay, so this is five.', 'start': 2690.216, 'duration': 2.182}, {'end': 2694.42, 'text': "so you can see, it's showing five.", 'start': 2692.398, 'duration': 2.022}], 'summary': 'Deep learning model representing numbers through patterns.', 'duration': 24.551, 'max_score': 2669.869, 'thumbnail': 'https://coursnap.oss-ap-southeast-1.aliyuncs.com/video-capture/sfheWK72L74/pics/sfheWK72L742669869.jpg'}, {'end': 2741.891, 'src': 'embed', 'start': 2713.452, 'weight': 1, 'content': [{'end': 2717.134, 'text': 'We basically are trying to create a bounding box and then give an output.', 'start': 2713.452, 'duration': 3.682}, {'end': 2724.217, 'text': 'This is a very simple example of what image processing is capable of all right guys with this we come to the end of our session.', 'start': 2717.654, 'duration': 6.563}, {'end': 2725.878, 'text': 'I hope you enjoyed and learn something new.', 'start': 2724.337, 'duration': 1.541}, {'end': 2730.019, 'text': 'If you have any further queries, please do mention them in a comment box below until next time.', 'start': 2726.158, 'duration': 3.861}, {'end': 2731.04, 'text': 'Goodbye and take care.', 'start': 2730.079, 'duration': 0.961}, {'end': 2733.928, 'text': 'I hope you have enjoyed listening to this video.', 'start': 2731.787, 'duration': 2.141}, {'end': 2741.891, 'text': 'Please be kind enough to like it and you can comment any of your doubts and queries and we will reply them at the earliest.', 'start': 2734.288, 'duration': 7.603}], 'summary': 'An introduction to creating bounding boxes in image processing.', 'duration': 28.439, 'max_score': 2713.452, 'thumbnail': 'https://coursnap.oss-ap-southeast-1.aliyuncs.com/video-capture/sfheWK72L74/pics/sfheWK72L742713452.jpg'}], 'start': 2053.614, 'title': 'Mouse button events and image processing in python', 'summary': 'Covers processing mouse button events, setting iswriting variable, appending coordinates to x and y chords and explains the process of image processing in python. it includes sorting x and y coordinates, finding and resizing images, drawing rectangles, and demonstrating a deep learning model recognizing handwritten digits.', 'chapters': [{'end': 2126.041, 'start': 2053.614, 'title': 'Mouse button events and coordinates processing', 'summary': 'Discusses processing mouse button events, setting iswriting variable, and appending coordinates to x and y chords, including the importance of the mouse button up event.', 'duration': 72.427, 'highlights': ['The importance of the mouse button up event and its role in performing various tasks after writing is done.', 'Setting the isWriting variable to true when a person starts writing.', 'Appending coordinates to x and y chords as part of processing mouse button events.']}, {'end': 2748.933, 'start': 2126.682, 'title': 'Image processing in python', 'summary': 'Explains the process of image processing in python, including sorting x and y coordinates, finding rectangles, resizing images, and drawing rectangles, with a demonstration of a simple example and a deep learning model recognizing handwritten digits.', 'duration': 622.251, 'highlights': ['The chapter explains the process of image processing in Python, including sorting X and Y coordinates, finding rectangles, resizing images, and drawing rectangles.', 'The demonstration of a simple example and a deep learning model recognizing handwritten digits is included.']}], 'duration': 695.319, 'thumbnail': 'https://coursnap.oss-ap-southeast-1.aliyuncs.com/video-capture/sfheWK72L74/pics/sfheWK72L742053614.jpg', 'highlights': ['The demonstration of a simple example and a deep learning model recognizing handwritten digits is included.', 'The chapter explains the process of image processing in Python, including sorting X and Y coordinates, finding rectangles, resizing images, and drawing rectangles.', 'The importance of the mouse button up event and its role in performing various tasks after writing is done.', 'Setting the isWriting variable to true when a person starts writing.', 'Appending coordinates to x and y chords as part of processing mouse button events.']}], 'highlights': ['Driverless cars are a notable application of image processing, with top players being Tesla, Audi, and Mercedes.', 'Image processing is used in medical diagnosis to compare MRI images and detect abnormalities using deep learning models.', 'The model consists of 1446 layers with a small number of parameters compared to other deep learning applications, as indicated by the model summary.', 'The process of creating a Convolutional Neural Network model using Keras framework is discussed, involving the use of convolution, filter size, input shape, and activation function to enhance non-linearity.', 'The chapter demonstrates training a model using Keras and saving it as an H5 file.', "Python's role in computer vision is discussed, highlighting its relevance and importance in the field."]}