- Slides
- Recording
- Project Tips:
- You will definitely need to use contours, color thresholding, and moments in your solution. Think about which order would be the most efficient.
- If you used the color picker linked in slides, you would have noticed that in some images, the red hue is between 0-5 and in other images it is between 165-180. Each of these will need their own mask, but think about how you can use cv2.bitwise_or to combine the masks together
- The saturation of the circle is higher than most of the other pixels in the image
- Use np.shape(img)[1] to get the width of the image and np.shape(img)[0] to get the height. Remember that the shape of images is (height, width, # color channels)
- Slides
- Recording
- "Part 1" is here, but there were technical issues and basically all of this is covered in Part 2, so it's not needed.
- Running Flask in Google Colab (for those of you with installation issues)
- Video
- Starter code for the project inside Colab (as opposed to just running it on your local machine)
- Neural Networks with Pytorch
- Docker + Unit Testing + CI
Kahoot questions covering OpenCV and Backend here.