Panoramic target track recording equipment and recording method

Abstract

The invention discloses a panoramic target track recording method, and the method comprises the following steps: enabling a main control module to perform target detection according to data of a panoramic image sensor array, and counting the number of detected targets; when the main control module detects that a new target is on line, calling a pan-tilt horizontal motor and a pan-tilt vertical motor according to the size of the target, and rotating the snapshot image sensor array to the corresponding position and angle; enabling the snapshot unit main control to collect the data obtained by the snapshot image sensor array in real time to snapshot a target, and obtaining a target ID; if the target does not exist in the local target library, warehousing and establishing a target ID; enablingthe main control module to update the target state according to the target ID; enabling the main control module to track the corresponding target ID; enabling the main control module to perform target state updating, acquiring the ID of the newly added offline target, generating a complete target trajectory from target entry to target loss, and storing the complete target trajectory. According tothe invention, intellectualization of video monitoring can be realized.

Description

technical field [0001] The invention relates to the field of panoramic target track recording equipment, in particular to a panoramic target track recording device and a recording method. Background technique [0002] At present, in security monitoring, long-term redundant recording is often carried out for specific scenes, resulting in a waste of video upload traffic and a waste of storage space for local historical videos. In order to solve this problem, it is a more effective solution to trigger recording of surveillance video. Common trigger solutions include motion detection and target detection. Due to the relatively high rate of false alarms and false negatives, the effect of motion detection is not ideal in actual use. Object detection can only simply extract the video during the time period when the target of interest appears in the monitoring field of view. For the same target, there are many The video clips that appear in the same scene for the first time will no...

AI Technical Summary

Problems solved by technology

Due to the relatively high rate of false alarms and false negatives, the effect of motion detection is not ideal in actual use. Object detection can only simply extract the video during the time period when the target of interest appears in the monitoring field of view. For the same target, there are many The video clips that appear in the same scene for the first time will not be merged, resulting in severe fragmentation of the generated video clips, and because the field of view of the monitoring equipment is often limited, the same target may enter and exit frequently in the field of view. It aggravates the fragmentation of the video clips of the same target; another serious defect is that due to the noisy environment in the surveillance environment, this type of equipment cannot record a clear sound source for the target in the outdoor scene; at the same time, due to the existence of surveillance cameras In the case of limited field of view, the target can easily move into the blind area of ​​the monitoring field of view in the monitoring scene, resulting in incomplete recorded target video tracks

In addition, in order to collect enough rich information, the target search box of ordinary target track recording equipment is often set relatively large, resulting in the target monitoring distance is relatively short, or the target search box is set relatively small, but because the effective resolution is not enough, it cannot Enough information is collected to extract the target features, even if the target trajectory is extracted, the target identity cannot be further judged

Can only get shallow information about people or cars passing by in the scene, and cannot locate the identity of the target

leading to its application is also subject to great limitations

Images