Object detection of colored images using improved point feature matching algorithm
Anuj Gupta, Manvinder Sharma, Dishant Khosla, V. P. Singh · 2019
For computer vision, image matching is an essential trait which includes scene or object recognition. Detection using point feature method is much effective technique to detect a specific target instead of other objects or within clutter scene in an image. It is done by comparing correspondence points and analyzing between cluttered scene image and a target object in image. This paper presents novel SURF algorithm that is used for extracting, describing and matching objects in colored images. The algorithm works on finding correspondence points between a target and reference images and detecting a particular object. Speeded-up robust features (SURF) algorithm is used in this study which can detect objects for unique feature matches and which has non-repeating patterns. This approach of detection can robustly find specified objects between colored cluttered images and provide constriction to other achieving near real time performance.