Missing child detection method based on human body detection

Abstract

The invention relates to a missing child detection method based on human body detection. In the case of very low power consumption, human body targets are initially detected using an infrared sensor. After a suspected target is found through initial detection, a camera is started, and whether there is a missing child is confirmed using a video target re-checking method. Through a double-detection mode, the power consumption of the system is reduced, the false alarm rate and missing alarm rate of missing child detection are reduced, and the reliability of missing child detection is improved.

Description

technical field [0001] The invention relates to a method for detecting missing children and belongs to the technical field of video surveillance. Background technique [0002] In recent years, there have been many child fatal accidents caused by missing children on school buses, which has aroused strong social concern. In order to prevent such accidents, in addition to regular safety education and supervision and inspection of the relevant responsible persons and children in the car, it is also necessary to improve the ability to discover and warn missing children technically. With the rapid development of computer vision technology, the human body detection technology is becoming more and more mature. The patent "Status Recognition and Danger Control System for Occupants Stuck in the Car" (No. CN201510779628.0, 2015) detects whether there are passengers on the seat through the pressure sensor Legacy, this detection method cannot distinguish between people and objects, nor ...

AI Technical Summary

Problems solved by technology

With the rapid development of computer vision technology, the human body detection technology is becoming more and more mature. The patent "Status Recognition and Danger Control System for Occupants Stuck in the Car" (No. CN201510779628.0, 2015) detects whether there are passengers on the seat through the pressure sensor Legacy, this detection method cannot distinguish between people and objects, nor can it detect children left behind in corridors or other areas where sensors are not installed, and the detection accuracy is not high

The document "Deeply Learned Attributes for Crowded Scene Understanding" (CVPR, 2015) uses a deep learning method to detect whether there is a human target in the image, but this method is not effective in detecting children with occlusions on the vehicle

The document "(Automatic Head Detection for Passenger Flow Analysis in Bus Surveillance Videos" (CISP, 2012) counts the number of people entering and exiting by setting up cameras at the door of the car. The counting accuracy is not high, and it is impossible to accurately determine whether children are missing from the bus.

Images