A Fast Image SIFT Feature Matching Method Based on GPU and Cascade Hash

Abstract

The invention discloses a fast image SIFT feature matching method based on GPU and cascaded hash, which belongs to the field of computer vision. According to the limitation of GPU global video memory size and memory size, the present invention establishes a three-level exchange mechanism of GPU, memory, and hard disk; at the same time, an improved GPU parallel protocol method is adopted to perform two hash maps with different encoding lengths on all SIFT feature points of the image ; proposed and used the upper triangular matrix block reading method; rough screening by local sensitive hash algorithm; fine screening of candidate points by calculating Hamming distance; finally, by calculating the Euclidean distance between the screening point and the point to be matched to find the most Appropriate matching point; using the asynchronous parallel method of CPU and GPU, the matching result data of the image pair is copied from the GPU memory to the memory and saved to the disk while the GPU is operating. The present invention greatly shortens the matching time of SIFT feature points, and can complete feature matching in a short time even in the case of massive data.

Description

technical field [0001] The invention belongs to the field of computer vision, and more specifically relates to a fast image SIFT feature matching method based on GPU and cascaded hashing. Background technique [0002] Image feature matching is the corresponding relationship of feature points between images, which has a wide range of applications in image recognition, image registration and 3D reconstruction. The SIFT feature has scale invariance and rotation invariance, so it has strong anti-interference ability and high matching accuracy, so it is a commonly used feature for image feature matching. However, each SIFT feature point is represented by a 128-dimensional vector. The similarity measure of the SIFT feature point is the Euclidean distance. In the case of a high-dimensional vector, the Euclidean distance calculation time is longer, and the linear search method matches between two images. The computational complexity is O(N 2 ), where N is the average number of fea...

Images