Picture processing based tunnel video moving object detection method

Abstract

The invention provides a picture processing tunnel video moving object detection method. The method includes the following steps of sequentially establishing an initial background model, a real-time dynamic update background model, a local structure similarity measure function, and a local gray level statistics measure function, and extracting a moving object area according to the local structure similarity measure function and the local gray level statistics measure function. According to the picture processing tunnel video moving object detection method, the real-time dynamic update background model is established, so that the influence of illumination variation on tunnel environment moving object detection can be effectively inhibited; the local structure similarity measure function and the local gray level statistics measure function are constructed, so that discrimination of moving objects and background information are improved, detection erroneous judgment brought by single structural information is reduced, the moving object area can be well partitioned, and accuracy of tunnel environment moving object detection is improved.

Description

technical field [0001] The invention relates to the technical field of tunnel traffic safety monitoring, in particular to an image processing-based tunnel video moving target detection method. Background technique [0002] With the rapid development of my country's economic and social construction, the number of car ownership has also increased step by step. The increase in car ownership has led to increased road traffic congestion, frequent traffic accidents, serious waste of energy, and deterioration of environmental quality. The emergence of these problems has put forward higher and higher requirements for smooth, safe and fast roads, and has also led to the large-scale construction and popularization of various high-level intelligent roads. Then my country's road system entered a period of great development and great leap. my country is a mountainous country, and tunnels have the advantages of improving the road network and saving land, so they have gradually become an ...

AI Technical Summary

Problems solved by technology

[0010] Fourth, the update rate parameter has a greater influence

If the update rate parameter of the mixed Gaussian model is assigned a small value, the entire Gaussian model cannot construct the background required by the current frame, resulting in longer background modeling time

If you assign a larger value to the update rate parameter, although the speed of building the model is accelerated, the Gaussian component that matches the current pixel always occupies a larger weight due to the use of a larger update rate parameter, thereby reducing the impact of the entire Gaussian model on the video. Suppression of various noises

Images