Image shielding method and system

Abstract

The invention relates to an image shielding method. The image shielding method comprises that when a monitoring platform receives a video obtaining request sent by a video access terminal, an integrated virtualization manager (IVM) forwards the video obtaining request to an intelligent video unit (IVU) / an intelligent envelop unit (IEU); the IVU / the IEU obtains corresponding original video information from the monitoring platform; the IVU / the IEU conducts extraction of background images and a foreground target to the original video information, and image semantic analysis of the foreground target is conducted; the IVU / the IEU enables obtained target semantic description information to be matched with preset shielding rules or a preset shielding standard model base, whether image shielding is required is identified, if the image shielding is required, shielding of partial areas or whole areas of the foreground target in the original video information is conducted through the IVU / the IEU, and video information after the image shielding is obtained; and the IVU / the IEU returns the video information after the image shielding to the video access terminal through the monitoring platform. According to the image shielding method and an image shielding system, intelligent object recognition technology is utilized to enable content in video images to be automatically analyzed and made out, the target is analyzed, then the image shielding is conducted, and therefore, workload of workers is reduced.

Description

technical field [0001] The invention relates to the technical field of intelligent image recognition processing applications, in particular to an image masking processing method and system. Background technique [0002] With the vigorous development of video surveillance application technology, the area and objects covered by video surveillance are becoming more and more extensive. The ensuing public privacy issues have also attracted people's attention. In order to avoid the public privacy content involved in the monitoring content, the current common method is to manually distinguish people or objects, and perform masking and other processing in the corresponding areas in the video. However, this processing process requires manual identification of a large number of surveillance videos, so there are many disadvantages, such as: 1) the workload is huge and the cost is high; 2) it cannot be processed once when the image is recorded, and it needs to be reprocessed; 3) Differ...

AI Technical Summary

Problems solved by technology

The ensuing public privacy issues have also attracted people's attention. In order to avoid the public privacy content involved in the monitoring content, the current common method is to manually distinguish people or objects, and perform masking and other processing in the corresponding areas in the video. However, this processing process requires manual identification of a large number of surveillance videos, so there are many disadvantages, such as: 1) the workload is huge and the cost is high; 2) it cannot be processed once when the image is recorded, and it needs to be reprocessed; 3) Different application requirements require reprocessing for different masking objects, which is time-consuming and labor-intensive, etc.

Images