Local Feature Acquisition Method of Multi-layer Vision Network Image Based on Virtual Reality

Abstract

AbstractAs an important part of image features, local image features reflect the changes of local information in the image, which are not easily disturbed by various changes such as noise, illumination, scale, and rotation. Faced with this situation, this paper proposes a method for local feature acquisition of multi-layer visual network images based on virtual reality. Based on virtual reality technology, the multi-layer visual network image is reconstructed in layers, and the reconstructed images are preprocessed by histogram equalization and denoising. The SUSAN algorithm and the SIFT algorithm are combined to realize the acquisition of local image features. The results show that compared with the original SUSAN algorithm and the original SIFT algorithm, the average running time of the researched method is shorter and the overlap error is smaller, which indicates that the researched method has lower time complexity and higher acquisition accuracy.KeywordsVirtual reality technologyMulti-layer visual network imageLayered reconfigurationPretreatmentSUSAN algorithmSIFT algorithmImage local feature acquisition