Remnant detection method of double-mask background update based on double background models

Abstract

The invention discloses a remnant detection method of double-mask background update based on double background models, and relates to the field of intelligent monitoring and computer vision. The method comprises the following steps: S1: reading in a monitoring video, and creating two background models; S2: detecting stationary foreground objects; S3: screening the detected stationary objects; S4: updating the double background models; and S5: marking detected remnants and outputting the same to video monitoring. According to the remnant detection method disclosed by the invention, the stationary objects are roughly detected by the comparison of the double background models, accurate remnants are obtained by further accurate screening, the remnant detection accuracy is further guaranteed on the updating method of two background models, and meanwhile the detection instantaneity is also guaranteed. Moreover, various public occasions are well applicable, and the interference generated by environmental changes (e.g., light, swing of objects blown by the wind) is effectively avoided.

Description

technical field [0001] The invention relates to the fields of intelligent video monitoring and computer vision, in particular to a remnant detection method based on a double background model and a double mask background update. Background technique [0002] Video surveillance systems are widely used in all walks of life, especially in residential areas, banks, supermarkets, subways, airports, museums, etc. Under normal circumstances, the above monitoring system is mainly composed of traditional closed-circuit television CCTV monitoring, which can record and store the monitoring scenes, but cannot send alarms to public security and other criminal acts in time, and requires a large number of staff to monitor the monitoring images at all times . Therefore, the CCTV-based video surveillance system can no longer meet the needs of modern people for security protection, so intelligent video surveillance has emerged as the times require, and will completely replace the former in th...

AI Technical Summary

Problems solved by technology

[0004] At present, the mixed Gaussian model is directly used to model the two backgrounds at the same time, which requires a huge amount of calculation, and it is difficult to meet the real-time requirements.

At present, the long stay of remnants can easily be updated into the background, resulting in the inability to detect remnants; at present, there is no strict elimination of interference in the detection of remnants, resulting in a very high false detection rate of remnants

Images