Weighted convolutional autoencoder-long short-term memory network-based crowd anomaly detection method

Abstract

The invention discloses a method for performing anomaly detection by a weighted convolutional autoencoder-long short-term memory network (WCAE-LSTM network). The method is devoted to perform anomaly detection and positioning by learning a generation model of a mobile pedestrian, thereby guaranteeing the public safety. The invention provides a novel double-channel framework, which learns generationmodes of an original data channel and a corresponding optical flow channel and reconstructs data by utilizing the WCAE-LSTM network, and performs the anomaly detection on the basis of a reconstruction error. In addition, for the problem of complex background, it is proposed that a sparse foreground and a low-rank background are separated by adopting modular robust principal component analysis decomposition; and a weighted Euclidean loss function is designed according to obtained background information, so that background noises are inhibited. The designed WCAE-LSTM network can not only perform the anomaly detection globally but also roughly locate an abnormal region locally; and through the joint consideration of global-local anomaly analysis and optical flow anomaly analysis results, finally robust and accurate detection of abnormal events is realized.

Description

technical field [0001] The invention belongs to the field of computer vision, in particular to the field of intelligent monitoring, in particular to a method for detecting crowd anomalies using a weighted convolutional self-encoding long-short-term memory network. Background technique [0002] With the development of the field of video surveillance, more and more surveillance cameras are used in public places to ensure public safety. However, there is not only too much redundant information in a large amount of surveillance video data, but also a severe challenge for surveillance personnel. Video analysis is time-consuming and tedious. Therefore, the development of a system that can automatically detect video anomalies plays a vital role in reducing human and financial resources. Public safety will be further assured if continuous or potential anomalies can be accurately detected. [0003] Most of the traditional video anomaly detection methods use optical flow information...

AI Technical Summary

Problems solved by technology

However, these generated models are susceptible to complex backgrounds and cannot reconstruct normal behavior well

Meanwhile, methods that detect anomalies based on the reconstruction error of the entire frame may ignore abnormal behaviors in a small region

In addition, Generative Adversarial Network (GAN) can also detect anomalies, however, GAN models are difficult to train and cannot capture the differences between moving objects well

In addition, the model is still unable to resolve the effects of complex backgrounds and small area anomalies

Images