Background fraud detection method and device and electronic equipment

Abstract

The invention discloses a background fraud detection method and device and electronic equipment. The method comprises the following steps: inputting a group of pictures to be detected into a pre-trained deep neural network to extract depth local features of each picture; calculating an interior point between the depth local features of every two pictures by using a random sampling consensus algorithm to obtain an interior point calculation result, wherein the inner points are used for representing that the depth local features of the pictures have consistency; constructing an adjacent matrix based on an interior point calculation result, wherein elements in the adjacent matrix are the number of interior points between the depth local features of the picture; clustering the adjacent matrixes by using an affinity diffusion clustering algorithm to obtain a plurality of clustering clusters, wherein each clustering cluster comprises a plurality of elements; determining whether background fraud exists in the group of pictures based on the plurality of clustering clusters. Picture background fraud can be effectively detected, and the possibility of false detection or missing detection isreduced.

Description

technical field [0001] The invention belongs to the technical field of network security, and in particular relates to a background-based fraud detection method, device and electronic device. Background technique [0002] In recent years, there have been a large number of cases of fraud using image background fraud on the Internet, which greatly damages the rights and interests of Internet users and financial institutions, and poses a threat to network security. [0003] In order to detect background fraud in pictures, there are currently methods proposed: using a content network to obtain a set of images to be detected with backgrounds removed and complementing the backgrounds; then, using a texture network to obtain digital features of the group of images ; Finally, use the Euclidean distance to measure the similarity of the group of images to detect whether there is image background fraud. [0004] However, in the above method, the effect of removing the background and th...

AI Technical Summary

Problems solved by technology

[0004] However, in the above method, the completion effect of removing the background and then supplementing the background depends on the content network used, and may lose some local features in the original image, resulting in false detection or missed detection.

In addition, the object in the actual picture to be detected may have multiple poses, which makes the operation of removing the background and complementing the background difficult to perform in actual operation, which may also lead to false detection or missed detection

Images