Image anomaly detection method based on deep convolutional generative adversarial network

Abstract

The invention discloses an image anomaly detection method based on a deep convolutional generative adversarial network. The method comprises the following steps: 1) performing data acquisition; 2) performing data processing; 3) constructing a network; 4) defining a loss function; 5) training the network; and 6) performing network verification. The image semantic context information extraction capability of the dilated convolution residual block and the image generation capability of the generative adversarial network are combined; and a multi-scale strip mask is designed and used to remove a partial region of the image to enhance the reconstruction effect. The distinguishability between the normal sample and the abnormal sample and the accuracy of detecting the abnormal sample and positioning the abnormal position are improved.

Description

technical field [0001] The present invention relates to the technical field of image anomaly detection, in particular to an image anomaly detection method based on a deep convolutional generative confrontation network. Background technique [0002] Image anomaly detection is a technique to detect whether there are abnormal samples in a given image and point out the location of the abnormal samples. Abnormal samples are a concept that exists widely in the real world. They often refer to samples that do not meet people's definition of normal patterns, such as defects in industrial products, abnormal symptoms in disease inspections, contraband in security inspections, and samples in surveillance videos. violations, etc. Therefore, image anomaly detection technology has great application prospects in safety detection, quality detection, medical diagnosis and treatment, and behavioral early warning. [0003] For anomaly detection problems, on the one hand, the frequency of anom...

AI Technical Summary

Problems solved by technology

In recent years, deep self-encoders have been widely used in self-reconstruction anomaly detection methods. However, due to the good generalization of deep self-encoders, the reconstructed image will be similar to the abnormal image, resulting in failure of anomaly detection.

If the generalization ability of the deep self-encoder is directly constrained, the reconstructed output image will be blurred and the error will become larger, which will not only fail to achieve accurate detection of abnormal samples, but also reduce the distinction between normal samples and abnormal samples.

Images