Event camera motion compensation method and device based on motion model

Abstract

The invention discloses an event camera motion compensation method and device based on a motion model. Comprising the following steps: sequentially reading event sequences output by an event camera according to a preset time window; converting the event sequence in the time window into an average time plane, and obtaining the centroid of the average time plane; performing convolution operation on the average time plane by adopting a Sobel operator; calculating to obtain horizontal and vertical gradients of an average time plane; updating parameters of the motion model by adopting a gradient descent method; and updating the position of the event by adopting the updated parameters of the motion model. According to the invention, the event position is updated by establishing the motion model, so that the effect of motion compensation is achieved, and motion compensation can be effectively performed on motion blur of a high-speed moving target caused by event accumulation.

Description

technical field [0001] The present application belongs to the technical field of target detection and tracking, and in particular relates to a motion compensation method and device for an event camera based on a motion model. Background technique [0002] When an ordinary camera takes an image, it shows the scene within the exposure time. When the object in the scene moves, the image of the scene will show the combination of all positions of the object within the exposure time. Due to the limitation of the frame rate of ordinary cameras, the high-speed moving object will appear blurred, which is called motion blur. Motion blur is the moving effect in the scene image, which is more obvious in the case of long exposure or high-speed moving objects in the scene. During object detection, motion blur can confuse the exact location of an object. [0003] Traditional frame-based cameras have limitations such as information redundancy, high-speed blurring, and low dynamic range. ...

AI Technical Summary

Problems solved by technology

Therefore, in the case of rapid scene changes, due to the compression of too many events, the generated time plane will be blurred to a certain extent, which will affect the characteristics of the target and the judgment of the target position, thereby affecting target tracking, obstacle avoidance, etc.

The specific performance is that the edge of the target in the time plane has a tail, which will cause certain troubles for the subsequent feature extraction and position detection.

Images