Set of mini projects I did using an OpenCV , an image processing library in python
Skills : Probablistic robotics - recursive state estimation using a parametric filter known as extended kalman filter , image processing using opencv
This deserves its own github page , hence here is the link : https://github.com/ayush-agarwal-0502/Kalman-filter-using-opencv
Skills : OpenCV
Used OpenCV skills to build a motion detector for a video camera . Took in 2 images at very small time gap , took absolute difference of them , performed image processing ( like converting to grayscale , gaussian blur , dilating , blurring ) on the difference image , then used find contours function to detect the moving objects (since the moving objects will be at different positions in the 2 images , the difference will be non zero ) , also found number of contours , thresholded the detected contours based on a certain area value , then drew a rectangle around the objects . then using the puttext function , wrote the number of objects detected , also wrote "something is moving " or "something is not moving " , and also an additional feature of giving warning if the moving object is close to the camera ( which would be when the area of the contour is very large compared to that of the screen )
(Would work better if this is deployed on cctv cameras as the moving objects will be smaller hence properly detected and enclosed in rectangles, in the picture , the hand moved more than the object (cube) hence got detected more.)
Skills : Haar Cascades , OpenCV
Haar cascades source : https://github.com/anaustinbeing/haar-cascade-files
Haar Cascades documentation : https://docs.opencv.org/4.5.3/db/d28/tutorial_cascade_classifier.html
Haar Cascades is a library in python which is used along with opencv to detect face , eyes ... anything you want (and its haar cascade is availaible) in the image or video . This library internally consists of trained machine learning models to detect the features .
Skills: OpenCV , PyTesseract
Pytesseract is a python library used to detect the words , characters in an image . I have printed the individual detected characters , as well as the complete words over the image .
