An Image Alignment Method for Video Surveillance and UAV

Abstract

The invention discloses a method for image alignment of video surveillance and unmanned aerial vehicles, comprising the following steps: step S1: aligning the coordinates of the monitor camera picture and the coordinates of latitude and longitude, and establishing a low-precision mapping relationship between the coordinates of the monitor camera picture and the coordinates of latitude and longitude; step S2: Align the UAV camera picture coordinates with the latitude and longitude coordinates, and establish a low-precision mapping relationship between the UAV camera picture coordinates and the longitude and latitude coordinates; Step S3: Screen the target objects corresponding to the positions in the surveillance camera picture and the UAV camera picture, and obtain mutual correspondence The longitude-latitude coordinate pair set K of the target object, the average value of each pair of longitude-latitude coordinates is taken as the standard longitude-latitude coordinate of this group, and the standard longitude-latitude coordinate set is obtained. The two transformation matrices are calculated, and the precise mapping relationship between video surveillance and UAV is established through the calculated transformation matrix.

Description

technical field [0001] The invention relates to the field of monitoring and detection, in particular to an image alignment method for video monitoring and unmanned aerial vehicles. Background technique [0002] Video surveillance and unmanned aerial vehicles are widely used in the ocean and land, and can conduct rapid daily monitoring of areas that are illegal or suspected of illegal use of the sea, illegal buildings and other characteristics, providing a basis for illegal use of sea investigations and emergencies. in accordance with. In order to quickly obtain the information of different angles of view of the target object, so as to play a better monitoring role, video surveillance and UAV image mapping are required. [0003] The existing technology generally adopts a method for image alignment to extract feature key points and feature descriptors of two images, and then use a matching algorithm to find feature points with a high matching degree for transformation to achi...

AI Technical Summary

Problems solved by technology

However, in this method, due to the fact that the manually calibrated coordinates contain errors in practical applications, the obtained image coordinates generally contain errors, which is time-consuming and labor-intensive.

Images