Automatic recording system for vehicles running red light

Abstract

The invention relates to an intelligent system integrating vehicle detection, image processing and photoelectric control, in particular to an automatic recording system for vehicles running a red light. The system comprises a vehicle detector, an image-collecting and shooting part, an auxiliary light automatic control part, a main control computer and a dynamic power supply. The invention has vehicle capture rate higher than or equal to 99 percent and accurate vehicle positioning, can record the whole process of vehicles running a red light and enable the acquired images to reach the resolution over 1624*1236 pixels, thereby being suitable for being used in all time and all weathers.

Description

technical field [0001] The invention relates to an intelligent system integrating vehicle detection, image processing and photoelectric control, in particular to an automatic recording system for vehicles running red lights. Background technique [0002] The existing automatic recording system records one close-up image of the front part of the passing vehicle and three panoramic images of the vehicle at the same time. The maximum resolution does not exceed 768*576 pixels. It is okay for daytime applications, but the nighttime effect will be poor. Many, the complete process of running a red light cannot be clearly recorded. [0003] There are three main reasons for the above results. First, the image acquisition method is unreasonable. The ordinary PAL camera scanning system adopts an interlaced scanning method, and the shutter is opened 50 times per second to generate 25 frames of images, and each frame of images is composed of odd numbers. Field and even field, the time d...

AI Technical Summary

Problems solved by technology

[0002] The existing automatic recording system records one close-up image of the front part of the passing vehicle and three panoramic images of the vehicle at the same time. The maximum resolution does not exceed 768*576 pixels. It is okay for daytime applications, but the nighttime effect will be poor. Many, cannot clearly record the complete process of the vehicle running the red light

[0003] There are three main reasons for the above results. First, the image acquisition method is unreasonable. The ordinary PAL camera scanning system adopts an interlaced scanning method, and the shutter is opened 50 times per second to generate 25 frames of images, and each frame of images is composed of odd numbers. Field and even field, the time difference between two adjacent images that make up one frame of image is about 20ms, calculated at a vehicle speed of 180 km / h, the vehicle in the two images will have a displacement of 1 meter, when watching the dynamic image, Due to the phenomenon of persistence of human vision, the image will be very smooth, but if the image is acquired by frame acquisition, an image composed of two fields will be obtained, and the vehicles in the image will appear ghosting or tailing, which is obviously not correct. It may be clear, but this phenomenon is more obvious if the vehicle moves laterally in the frame

The second is that the response speed of the vehicle detector is slow, and the synchronization speed of the image acquisition card is slow. As a result, when the system collects pictures, the license plate part has left the camera shooting area, and empty shots occur. Sometimes the vehicle in the panoramic view and the vehicle captured by the close-up camera are inconsistent. caused by this situation

The third is that the backlight compensation and strong light suppression functions of the camera take a long time. Due to the high speed of the vehicle during the system shooting, the vehicle has passed before the camera can reach the high speed. At this time, the captured image is either too dark or too bright. good for vehicle photography