Neural Network Based Implementation of Corner Detection for Biomedical Application in Computer Vision

Abstract

The identification of the corner is one of the important features of the visual processing system. When the
image is extracted and used to modify the image, image may be enhanced according to the application, such
as corner detection, edge detection, sharpness, etc. Different devices, such as real-time video surveillance,
medical imaging, facial recognition, include object detection, including detection in a corner. We are
proposing a Corner Detection Method with Neural Network in order to provide improvement over existing
corner detector method. The theoretical rigor benefits are there in conventional differential filter-based
algorithms but excessive post-processing is required. The proposed new network technique is used in the
realization of corner detection tasks, it takes advantage of the extraction of momentum, can process an input
picture of any size without preparation.