This is a real-time object recognition project developed using Python Programming Language
- First, it starts off the webcam for video capture
- Now we extract frame by frame from the image and do analysis on it
- (Note we are analyzing frames in grayscale)
- Now we have a loaded training image that will be used to track
- Since we got the tracking image now we will compare it with each frame we will get from the video
- As the pixels of tracking photo matches with certain pixels on the frame, we will draw a rectangle over there at ROI (Region of Interest)
- Hence as soon as we got a frame and object is recognized there will be a new rectangle formed.
- Add your own image into the images folder
- Name your image like "pic" + number + ".png"
- And give reference of it into the python code
- Change Number of images in the code & Done!
- Run the code :)
- Track as many as images at a same time
- Change the threshold as on your own requirement as see the result