Using OpenCV I processed any given sudoku puzzle, using this process
- Converting it to GrayScale
- Gaussian Blurring the Image
- Adaptive Thresholding to contrast the image
- Finding the contour to find the corners of the Sudoku
- Warped the perspective of an image to give just the Sudoku grid
- Using a created split_Boxes() method to divide the image into 81 even images
- Using CNN to detect the numbers and place them into an array
- Overlay the array onto a blank image
- Using a created drawGrid() method to draw a grid around each number
- Warp the perspective back to how the photo is originally given
- Overlay the numbers onto the original image