DetectBaseballPitch

An algorithm that can take any pitch from the camera and detect the frame where the ball was released from the pitcher’s hand.

This mini-project aims at recieving a video input from the videos of Matt Harvey pitching a baseball. Using Python and OpenCV techniques the user can determine in what frame Matt Harvey pitched the ball.

Requirements:

Python 2.x , 3.x

OpenCV

Techniques used:

1. Video breakdown into frames
2. Worked with a single image channel and cropping to reduce processing time of 4K images
3. Gaussian Blur to denoise
4. Canny edge detector and alongwith dilation and erosion techniques
5. Finding contours in each frame and the isolating the ball from pitcher after it is released from his hand
6. Sorting the contours found and returning the frame with a contour area <100 (The obserevd area of the ball/circle)

test