Jaywalking snapshot method and device

Abstract

The invention belongs to the field of video monitoring, and discloses a jaywalking snapshot method and a jaywalking snapshot device. The method comprises the following steps of: sequentially shooting high-definition pictures, performing jaywalking identification on the contents of the high-definition pictures and computing the speed of a vehicle through a high-definition camera and an embedded digital signal processor; when an effective jaywalking event is determined, combining three high-definition pictures into one jaywalking picture; and transmitting the picture to a remote-end server through a network. Due to the adoption of the method and the device, false shooting and leaked shooting can be effectively prevented; particularly, a number plate and the states of traffic lights can be clearly distinguished; and the load of the network can be further reduced obviously, and remote-end servers and database resources can be saved.

Description

technical field [0001] The invention belongs to the field of video surveillance, and more specifically, relates to a method and device for snapping photos of running a red light. Background technique [0002] Most of the existing red light capture devices are standard-definition cameras. Because the standard resolution of standard-definition cameras is low, the field of view is not wide enough, and the coverage is narrow, it is difficult to cover multiple lanes, so multiple cameras are needed to cover multiple lanes, resulting in Waste of resources. like figure 1 As shown, the lanes L1, L2, L3, and L4 are respectively equipped with a standard definition camera C1, C2, C3, and C4. In addition, the pictures usually taken by standard definition cameras are often unable to see the whole picture of the vehicle, clearly identify the license plate, or provide any correct, complete and clear vehicle information due to the low resolution. Therefore, the pictures after shooting mu...

AI Technical Summary

Problems solved by technology

In addition, the pictures usually taken by standard definition cameras are often unable to see the whole picture of the vehicle, clearly identify the license plate, or provide any correct, complete and clear vehicle information due to the low resolution. Therefore, the pictures after shooting must be discarded, resulting in A large increase in the amount of discarded data and unnecessary waste of resources

On the other hand, there is not enough ideal identification method and strict identification process for running a red light, which often results in wrong or missed shots, and wastes system resources

In addition, although some more advanced red light capture devices have begun to use high-definition cameras in an attempt to solve the defects of standard-definition cameras, because the capture of the near-end high-definition camera is controlled in the remote server, the captured high-definition pictures are transmitted to the remote end through the network The server will then perform red light identification

However, high-definition pictures have a large amount of data, and transferring them to a remote server through the network will seriously occupy bandwidth, and the effect is extremely unsatisfactory

Moreover, the process of running a red light and the identification method have not been improved. The high-definition pictures transmitted for each snapshot cannot ensure that they are correct photos of running a red light. Mistakes and missed shots still exist, and the high-definition pictures are even bigger. End server and network load

Images