Feature Detection

Learn to combine feature matching and findHomography from the calib3d module to find known objects in complex images. We cover RANSAC-based robust perspective transform estimation.

Learn how to match features in one image with others. We cover Brute-Force matcher and FLANN Matcher with ORB and SIFT descriptors in OpenCV.

Learn about ORB, a free alternative to SIFT or SURF. We cover cv.ORB_create() for keypoint detection and descriptor computation using FAST + rBRIEF combination.

Learn the basics of BRIEF algorithm, a fast binary descriptor that uses only 32 bytes compared to SIFT's 128 dimensions, enabling high-speed matching with Hamming distance.

Learn about FAST (Features from Accelerated Segment Test) algorithm designed for real-time applications. We cover cv.FastFeatureDetector_create() with and without non-maximal suppression.

Learn the basics of SURF algorithm, a speeded-up version of SIFT that uses box filters and integral images to achieve 3x speedup while maintaining comparable performance.

Learn about the concepts of SIFT algorithm. We cover scale-space extrema detection, keypoint localization, orientation assignment, and finding SIFT keypoints with OpenCV.

Learn about Shi-Tomasi Corner Detector, an improvement over Harris. We cover cv.goodFeaturesToTrack() to find the N strongest corners in an image.

Learn about Harris Corner Detection concepts and implementation. We cover cv.cornerHarris() and cv.cornerSubPix() for sub-pixel accuracy corner detection.