A method, device and medium for intelligent monitoring and identification of the surrounding environment of a cashier

Abstract

The invention discloses a method, device and medium for intelligent monitoring and identification of the surrounding environment of a cashier. The method includes video monitoring, target tracking, interactive behavior detection and identification; Markov Monte Carlo uses the human body in the surveillance video as the target for multi-target tracking to obtain a tracking video including multiple target tracking trajectories; interactive behavior detection is: extracting features from the tracking video and constructing combined features to pass through the convolutional neural network The algorithm judges the facial avatar and its feature points of the human target, and then uses it to judge whether the head posture and gaze angle are abnormal, and detects the interaction behavior if it is abnormal; the interaction behavior recognition is: if there is an interaction behavior, further Identify the type of interaction. The invention can monitor people and objects in the withdrawal area in real time, and automatically detect and identify people's behavior.

Description

technical field [0001] The invention belongs to the technical field of behavior recognition, and in particular relates to a method, device and medium for intelligent monitoring and recognition of a cashier's surrounding environment. Background technique [0002] With the popularization of payment methods such as Alipay and WeChat, small amounts are less and less carried out in bank ATMs, and only some transactions with large amounts are easy to use cash. And most of the people who use cash transactions are older people. Therefore, it is necessary to focus on the safety of withdrawals for this group of people. [0003] At present, the monitoring of the surrounding environment of the cash machine mainly relies on the camera installed on the ceiling and the pinhole camera on the ATM for real-time monitoring and video recording. It is manually interpreted during working hours, but the ATM is open 24 hours, so the video is recorded after get off work, and Save the whole day vide...

AI Technical Summary

Problems solved by technology

However, it is not possible to warn the cashier so as to avoid accidents at the first time. In most cases, it is even due to the loose staff and lack of background management.

[0004] The existing technology uses manual interpretation, but because the time spent staring at the background monitoring at work is very long, the length of video recording is often very large and the number of storage days is limited, the accuracy of interpretation depends entirely on the operator's visual recognition, which is difficult to guarantee, and the interpretation results have human subjectivity

At present, there is no intelligent video surveillance equipment for financial withdrawals in China

Images