OpenCV is a huge open-source library widely used in computer vision, artificial intelligence and image processing domains. The quintessential applications of it in real world are face recognition, object detection, human activity recognition, object tracking and such others.

Now, imagine you have to only detect an object from an entire input frame. So, instead of handling an entire frame, what if you can define a sub-region in the frame and treat it as the new frame to apply the processing. How do we do that you ask? Well, we are going to have a hands-on experience of implementing that in three different stages:

• Defining Region of Interest(ROI)
• Detecting Contours in ROI
• Thresholding the frame outlined by detected contours

Let’s get started, shall we?

#opencv #computer-vision #contour #threshold #hand-tracking