Method and system for detecting residues in atm panel area

Abstract

The invention discloses a method and system for detecting residues in the panel area of ​​an ATM machine. The method includes: step S1, acquiring a video stream in the area of ​​the ATM machine, processing the obtained video stream by using a background modeling method, and detecting the video stream Pedestrians entering or leaving; step S2, intercepting a frame of image respectively when pedestrians entering and pedestrians leaving are detected, and performing differential processing on the pedestrian entering image and the pedestrian leaving image, and extracting the gray scale change area of ​​the image; step S3, Segment the image of pedestrians leaving the image through a trained convolutional neural network model to extract the suspected remnant area; step S4, perform overlap calculations on the grayscale change area extracted in step S2 and the suspected remnant area extracted in step S3, In order to determine whether there are residues on the panel, the present invention can still accurately detect the residues in the panel area even in the case of environmental changes such as interference from light conditions.

Description

technical field [0001] The invention relates to the technical fields of image processing and video intelligent monitoring, in particular to a method and system for detecting residues in an ATM machine panel area based on a deep convolutional neural network. Background technique [0002] At present, automatic teller machines (Automatic teller machines, ATMs) have become an indispensable part of people's lives, and each bank has a large number of ATMs. However, while ATM brings convenience to people's life, it also brings some troubles, such as forgotten items and so on. Usually, banks will install cameras above ATMs to monitor financial crimes while also monitoring remnants. Therefore, the detection and analysis of leftovers on the panel has become one of the important components of the ATM security system. Its main function is to judge whether there is any leftovers on the panel area of ​​the ATM after the user leaves the ATM through a series of algorithm judgments. Items ...

AI Technical Summary

Problems solved by technology

[0003] At present, the existing ATM machine panel residue detection algorithm is based on traditional digital image analysis technology, which mainly judges by comparing the gray level of the panel images before and after pedestrians enter and exit, and by judging the gray level change degree of the image in the panel area Whether there are leftover objects, this method is simple and direct, but because this method is extremely sensitive to changes in the image area, there are the following problems: 1. Misjudgments may be caused by environmental changes or other factors, such as indoor and outdoor environments, lighting Changes in intensity and direction, and irregular entry and exit of customers may cause false alarms for remnants; 2. When the leftover objects "disappear" from the panel area, it will also trigger an alarm for remnants

Images