IDENTIFYING COMPLEMENTARY CORNER DETECTORS FOR CORRECT IMAGE PIXELS CLASSIFICATION

Classification of digital image content is mainly done by identifying low level imagefeatures such as corners and edges. The literature shows variety algorithms for the identification ofcorner non-corner pixels, important objects’ segmentation. However,all these produce different results same data therefore, suitable limitedapplications. This paper proposes a hybrid solution combining complementary corner detectionalgorithms to improve pixels’ classification. has been bestdetection algorithm points with small large angles producing bycombining latter two. Results have shown that Harris detector combined Global LocalCurvature Points (GLC) improved detection rate 28% in synthetic images, but 50% realimages whereas, combination Shi’s GLC enhanced rateby 25.9% images 123% real showing significant improvement.