Feature point matching method and electronic device based on event frame

Abstract

The present application provides an event frame-based feature point matching method and electronic device. The method is applied in an electronic device containing an event camera. In this method, the event camera outputs an event stream and accumulates a certain number of event streams to generate event frames. After the electronic device obtains the first event frame and the second event frame and forms a pair of event frames, it extracts feature points from the first event frame and the second event frame respectively, and compares all the feature points in the first event frame with the second event frame. Perform point set matching on all feature points in to obtain the coarse affine transformation parameter P1. Then, the coarse matching points corresponding to each feature point in the first event frame in the second event frame are obtained according to P1. After that, template matching is performed between the neighborhood of each feature point and the corresponding coarse matching point neighborhood to obtain the fine affine transformation parameter P2. In addition, erroneous matching point pairs in the set of precise matching point pairs obtained based on P2 can also be filtered out, so as to obtain a final set of feature point matching pairs.

Description

technical field [0001] The present application relates to the field of image processing, and in particular, to an event frame-based feature point matching method and electronic device. Background technique [0002] With the rapid development of sensor technology and computer vision technology, augmented reality (AR), virtual reality (VR) and autonomous driving are increasingly being applied. Structure from motion (SFM) and Simultaneous localization and mapping (SLAM) play key roles as core technologies in AR, VR, and autonomous driving. Technologies such as SFM and SLAM can determine the motion trajectory of the camera through the observation of the environment, and at the same time construct a map of the environment. Among them, it specifically involves extracting feature points from images, matching feature points between different images, and then performing subsequent operations on the matched feature point pairs, such as estimating the motion trajectory of the camera t...

AI Technical Summary

Problems solved by technology

At present, when using standard cameras to capture images, there are problems such as motion blur, overexplosion, underexplosion, and high time delay, resulting in poor output image frames, which will further lead to great difficulties in feature point extraction and feature point matching.

Images