Multi-camera crowd-gathered message statistic device and method

Abstract

The invention discloses multi-camera crowd-gathered message statistic device and method. The method comprises the following steps of: obtaining visible images and infrared images of human objects in a monitored region by at least three monitoring camera subassemblies arranged in different monitoring positions of a monitored region, wherein each monitoring camera subassembly comprises a visible camera, an infrared camera and a bracket; transmitting the visible images and the infrared images into a computer by a video collecting and coding device to register and inosculate in order to form a combination image; identifying the mode of the combination image by the computer to obtain a human object position and a gathering message; eliminating a forged human object according to the object thermal radiation condition in the infrared images; eliminating identification errors caused by sheltering and forged objects by the computer which carries out arbitration treatment aiming at different human object position messages of the combination images obtained by the camera subassemblies after mode identification; proving geometric parameters and optical parameters of the camera subassemblies by a proving device; and alarming by an alarm device when the crowd gathering is ensured to be detected.

Description

technical field [0001] The invention belongs to the field of computer vision, and in particular relates to a device and method for reliably calculating crowd gathering information in a monitored area in security monitoring. Background technique [0002] In the field of security monitoring, especially in the supervision of prisoners in prisons and the identification of traffic vehicles, it is often necessary to count, track and even identify whether there is crowd gathering in the monitored area. [0003] The traditional monitoring method is manual monitoring and counting, and security personnel need to be placed in the monitored area, which has low efficiency, poor accuracy, and cannot achieve objective judgment, so it can only be used for a short time and in a small area. [0004] In order to overcome the deficiencies in traditional manual monitoring and counting, video monitoring technology has emerged at home and abroad, that is, the camera installed in the monitored area...

AI Technical Summary

Problems solved by technology

However, the above-mentioned intelligent video monitoring method based on pattern recognition technology cannot overcome the following difficulties:

[0007] (1) When the video surveillance target is occluded, especially when it is occluded for a long time, the reliable positioning of the target cannot be achieved;

[0008] (2) In the field of security, especially in the security monitoring of key areas, there may be false targets such as dummies and pictures, and this method cannot realize the removal of the false and the preservation of the true;

[0009] (3) At night or in low visibility conditions, the above methods cannot achieve reliable monitoring

Images