
Research on Image Recognition Method Based on Deep Learning
Author(s) -
Jiebo Luo
Publication year - 2019
Publication title -
journal of physics. conference series
Language(s) - English
Resource type - Journals
SCImago Journal Rank - 0.21
H-Index - 85
eISSN - 1742-6596
pISSN - 1742-6588
DOI - 10.1088/1742-6596/1395/1/012008
Subject(s) - artificial intelligence , computer science , preprocessor , pattern recognition (psychology) , image (mathematics) , computer vision , feature extraction , artificial neural network , image processing , convolutional neural network , feature (linguistics) , convolution (computer science) , deep learning , identification (biology) , philosophy , linguistics , botany , biology
With the rise of artificial intelligence technology, computer vision, image processing, image recognition, voice recognition and other technologies have gradually entered the public’s vision. This paper focuses on image recognition. This paper discusses the original image acquisition, image information preprocessing, image feature extraction, pattern classification and recognition results. Analyzes VGG-19 network, VGG improved has realized the image quick recognition, studies input and output points, the number of neurons by convolution neural network, and finally concludes that the deep learning method can shorten the training time, improve the identification results, also the requirement for images in the image library is greatly reduced, the demand for hardware is not too high.