Vehicle-behavior analysis and identification method based on holder and camera device

Abstract

The invention discloses a vehicle-behavior analysis and identification method based on a holder and a camera device. The method includes the following steps: performing movement foreground detection through a mixed Gaussian background modeling method based on image weak edges; detecting movement foreground areas through use of a vehicle detection algorithm based on cascaded LBP space histogram features and judging whether vehicles are available in the areas and locating the positions of the vehicles accurately; performing matching and locating a target vehicle for a long time through use of a rapid vehicle matching algorithm based on class-Fern features; judging whether the target vehicle is illegally parked through use of the number of matched features of the target and after a preset position of the holder is switched, analyzing and identifying a before-event target vehicle; and realizing all functions of acquisition of real-time video images, vehicle-behavior analysis, holder linkage, image compression and network transmission and the like through use of an insert-type integral hardware structure. Device management roles of enabling according to time periods and enabling according to preset positions are also played. Rear-end platform software is only needed to receive pictures and to execute data compilation and selection work.

Description

technical field [0001] The invention belongs to the technical field of intelligent transportation, and relates to a vehicle behavior analysis and recognition method based on a cloud platform and a camera device. Background technique [0002] In the intelligent transportation industry, surveillance cameras and PTZs have become very popular, but the intelligence level of vehicle behavior detection and analysis is relatively low, and the ability to automatically detect violations (violation parking, retrograde, etc.) is not strong, and most of them send special personnel to check , so not only need to invest a lot of human and material resources, but also due to the limited energy and attention of people, in a high-intensity working environment, omissions often occur, so that it is impossible to respond to abnormal situations in a timely manner. Contents of the invention [0003] In order to solve the above-mentioned problems in the prior art, the present invention provides a...

AI Technical Summary

Problems solved by technology

[0002] In the intelligent transportation industry, surveillance cameras and PTZs have become very popular, but the intelligence level of vehicle behavior detection and analysis is relatively low, and the ability to automatically detect violations (violation parking, retrograde, etc.) is not strong, and most of them send special personnel to check , this not only needs to invest a lot of human and material resources, but also due to the limited energy and attention of people, in a high-intensity working environment, omissions often occur, so that it is impossible to respond to abnormal situations in a timely manner

Images