An Intelligent Supervision System for Fish Video Recognition

Abstract

The invention provides an intelligent supervision system for fish video recognition, including a business layer, a data layer, and a presentation layer. The business layer includes three major systems: a user system, a business system, and an algorithm system. The user system is mainly used to manage platform user operation behaviors and information management; the business system is used to manage the business between the entire platform modules; the algorithm system realizes the video recognition of fish through the fish model clustering algorithm. In the present invention, by fixing the fish model at the origin of the coordinate system, aiming at the characteristics of color clustering and fish video, an independent fish clustering color library is constructed by using the gray distribution statistics of fish schools and background objects; The clustering time is reduced and the efficiency is improved; the clustering value is determined adaptively through the valley value of the normalized histogram, which avoids inefficient and ineffective segmentation caused by manual experience.

Description

technical field [0001] The invention belongs to the field of intelligent supervision systems, in particular to an intelligent supervision system for fish video recognition. Background technique [0002] At present, big data has become a basic and strategic resource for national economic and social development. With the continuous development of information collection technology, various basic data such as regional resource and environment background data, management business data, and monitoring and monitoring data have grown rapidly, gradually showing the characteristics of big data. Management departments at all levels have increasingly strong demands for real-time, intuitive display and analysis of big data. It also puts forward more comprehensive and urgent requirements for the expansion, mining and application of spatial resource big data. At present, various business management systems are independent of each other, there is a lack of integration and sharing of infor...

AI Technical Summary

Problems solved by technology

At present, various business management systems are independent of each other, there is a lack of integration and sharing of information resources, the phenomenon of "information islands" is serious, deep data applications are insufficient, and data update mechanisms are not perfect.

[0003] Fish classification has been paid more and more attention for decades, and it is very important to rationally regulate fish populations. Effectively and accurately extracting and classifying fish is very important for analyzing and identifying the relationship between fish behavior characteristics and environmental factors and accurately It is essential to control the growth environment of fish, but it is currently based on manual identification, and there is a lack of effective algorithmic identification methods for fish videos

Images