Freshwater fish image real-time identification method based on lightweight convolutional network

Abstract

The invention discloses a freshwater fish image real-time identification method based on a lightweight convolutional network, and aims to solve the problems of lack of data sets, difficulty in feature extraction and low identification speed in the existing freshwater fish species classification technology. The method comprises the following specific steps: (1) constructing a freshwater fish image data set; (2) constructing a lightweight deep convolutional neural network; (3) training a deep convolutional neural network; (4) identifying the freshwater fish test image in real time; by constructing the freshwater fish image data set and utilizing the trained lightweight deep convolutional neural network, real-time recognition is automatically performed on the collected freshwater fish image, and the method has the advantages of no need of manually extracting fish body features, high recognition precision, high speed and small hardware resource consumption.

Description

Technical field [0001] The present invention belongs to the field of image processing technology, and more, the real-time identification method of a freshwater fish image based on a lightweight convolutionary network in the field of image recognition. The present invention can be used in fishery monitoring, aquaculture, and leisure scenes for real-time monitoring and identification of captured freshwater fish. The identification results can be used for fish species collection, and can provide reference for rare fish species. Background technique [0002] During the traditional fish identification process, it is mainly used to identify the fish, and the automatic freshwater fish recognition technology can significantly reduce the intensity of artificial labor, which can be used to fisheries monitoring, aquaculture, etc. In recent years, the machine learning method based on image characteristics has been applied to fish image recognition, and it has achieved good results. However, ...

AI Technical Summary

Problems solved by technology

[0005] The purpose of the present invention is to address the deficiencies in the prior art above, and propose a real-time recognition method for freshwater fish images based on lightweight convolutional networks, aiming to solve the problems of lack of data sets and difficult feature extraction in the existing freshwater fish recognition technology. , the problem of slow recognition speed

Images