Motion estimation method, electronic equipment and storage medium of a spatio-temporal pulse array

Abstract

The invention provides a motion estimation method of a space-time pulse array. The method comprises the following steps: dividing the space-time pulse array into coding cubes; determining a search range and a search starting point of the time domain and the space domain according to the feature information of the current coding cube; projecting all pulse signals in the space-time pulse array on aspace plane, and then carrying out distance measurement and matching analysis; judging whether to terminate the search in advance or not through a matching threshold of the current coding cube and the reference cube, if so, terminating the search in advance, otherwise, continuing the search until the search range is exceeded; screening an optimal reference cube; and outputting the motion estimation coding information of the spatio-temporal pulse array. According to the method, the high-precision motion vector can be effectively acquired, and meanwhile, the motion estimation search process isaccelerated, so that the problems of limited coding performance and long time consumption of motion estimation in a space-time pulse array coding signal are solved.

Description

technical field [0001] The invention relates to the technical field of information coding, in particular to a motion estimation method for a space-time pulse array. Background technique [0002] In recent years, spatio-temporal pulse array signal data is ubiquitous in the fields of computational neuroscience, computer vision, social media, earth and atmospheric science, and transportation, and emerges in the form of massive data. [0003] The space-time pulse array signal is a sparse space-time lattice described by the space-time point process model, which is abstract, discrete and hierarchical. The abstraction of spatio-temporal point process is a set of abstract discrete point events on different time scales and spatial scales, for example, nerve pulses, point clouds, social media user points, earthquake event points and traffic incident points, etc.; the discreteness of pulse array signals is Sparse lattice with discontinuous distribution in time domain and space domain;...

AI Technical Summary

Problems solved by technology

[0007] The present invention provides a motion estimation method of spatio-temporal pulse array, which can effectively obtain high-precision motion vectors, and at the same time accelerate the search process of motion estimation, so as to solve the coding performance limitation and time-consuming time of motion estimation in spatio-temporal pulse array coded signals The problem

Images