Year: 2019
Type: conference-paper
Abstract: Due to the impact of global environmental pollution deterioration, as well as uncontrolled fishing activities. The global fish stocks are becoming more and more serious, so it is necessary to pay attention to this issue. In this paper, the characteristics of traditional fish identification methods are vast, subjective factors are strong, and the current accuracy is not good in today’s deep learning. A migration learning based on Inception_v3 network with excellent feature extraction ability is proposed. Fish recognition model This model has better recognition ability than traditional models and some simple convolutional neural networks, and further improves the accuracy of Inception_v3 for fish recognition.
Authors: Jiangchang Huang, Weijie Gong, Hong Chen
Institution: Shenzhen University
Language: English
Subfield: Computer Vision and Pattern Recognition
Field: Computer Science
Domain: Physical Sciences
SDG: Life below water
Open Access status: diamond
OpenAlex ID: W2994983947