Small target intelligent identification system based on remote video monitoring and identification method thereof

Abstract

The invention discloses a small target intelligent identification system based on remote video monitoring. The method is realized based on a remote video monitoring system. The remote video monitoringsystem comprises a front-end unit and a rear-end unit, the front-end unit is designed to have the functions of monitoring acquisition, intelligent analysis and a servo system; the rear-end unit is designed to have the functions of system control, multi-system integration and big data analysis, and the small target intelligent identification system is located in the front-end unit and comprises avideo acquisition module, an image preprocessing module, an intelligent analysis module, a data processing module, a coding analysis module and a data communication module. Compared with the prior art, the system can effectively realize intelligent, accurate and real-time detection and identification of small targets under the condition of remote video monitoring, and transmits image identification information and derived early warning information to a rear-end control center.

Description

technical field [0001] The invention relates to the technical fields of target detection, computer vision and video monitoring, in particular to a small target intelligent recognition system and a recognition method based on remote video monitoring. Background technique [0002] Remote video surveillance systems are more and more used in border surveillance, sea area surveillance, remote observation and other fields, and become the main components of surveillance, reconnaissance and defensive systems. With the comprehensive application of visible light, infrared, radar, laser and other detection technologies, a full-time and all-weather video surveillance network has been established. At this stage, the core of video surveillance lies in intelligent, rapid and accurate identification of targets. [0003] In the practical application of remote video surveillance, the scenes are characterized by diversification and complexity, and are greatly affected by the environment. At ...

AI Technical Summary

Problems solved by technology

At the same time, there are still problems such as small scene target images, low resolution, and high background fusion, which brings great difficulties to remote target detection and recognition.

There are many methods for target recognition. The commonly used method is to establish a target model based on image processing and algorithm extraction of target features, and then perform target recognition based on deep learning technology. This method is easily affected by the external environment and real-time Poor, poor target recognition rate

Images