ESDF map construction method based on light field imaging

Abstract

The invention relates to an ESDF map construction method based on light field imaging, the light field imaging is to simulate a binocular camera by adding a micro-lens array between a main lens and a sensor plane, and due to the existence of the micro-lens array, the spatial information and angle information of light are obtained at the same time; and the angle information is integrated to form depth information of each pixel point. And meanwhile, the pictures shot by the light field camera cannot cause barrel shape distortion. Therefore, the method has the advantages that the depth map can be generated according to light field imaging, the occupied space on the unmanned aerial vehicle is relatively low, barrel distortion cannot be generated and the like to replace application of a binocular camera in SLAM is replaced.

Description

technical field [0001] The invention belongs to image guidance and relates to an ESDF map construction method based on light field imaging. Background technique [0002] When the accuracy of the traditional navigation method is insufficient or fails, the visual navigation system with rich and accurate information, less interference and strong real-time performance is usually used as an auxiliary positioning method. It is the current mainstream technology to use the data of visual sensors to construct maps of unknown environments in real time to realize the autonomous navigation of UAVs in unfamiliar scenes. There are three common map forms: geometric maps, grid maps and point cloud maps. The point cloud map has the advantages of being able to reproduce the unknown environment as it is, and the generation method is simple, but because it contains so much information that the large amount of point cloud map data is not conducive to preservation, so grids are often used in real...

AI Technical Summary

Problems solved by technology

[0003] Visual SLAM is divided into three types: monocular, binocular and RGB-D according to the sensor type. Because binocular cameras and RGB-D cameras can obtain the depth information of the scene, they are widely used in SLAM. However, due to the existing binocular The camera is large in size and heavy in weight, and the binocular camera will produce barrel-shaped distortion when acquiring images, making the subsequent image processing steps more cumbersome, so there are still problems in autonomous flights that require light weight and high speed. room for improvement

Images