Deep learning-based fancy carp screening method and device

Abstract

The present invention discloses a deep learning-based fancy carp screening method and a device. The method comprises the steps of collecting fancy carp images of a preset number, and subjecting each image to the centralized, size-standardized and normalized pre-treatment so as to obtain the feature vectors of the image; subjecting a deep belief network model to unsupervised pre-training and supervised fine tuning by utilizing a training sample, and outputting the extracted feature vectors of the training sample after being trained; adopting the output of the trained deep belief network as the input of a support vector machine classifier, and training the classifier so as to obtain corresponding parameters; and classifying the fancy carp by utilizing the trained classifier. According to the technical scheme of the invention, the deep belief network model for training a large amount of fancy carp image data is adopted and applied to the screening process for the high-quality seedlings of the fancy carp. Therefore, the dependence of the labor operation and breeding enterprises on the professional technical personnel is greatly reduced. The screening precision and the screening efficiency are improved.

Description

technical field [0001] The invention belongs to the technical field of computational pattern recognition, in particular to a koi screening method and device based on deep learning. Background technique [0002] Deep learning is a new research field in machine learning theory and an extension of the field of artificial intelligence. Its motivation lies in establishing and simulating the mechanism of the human brain to interpret data and analyze and learn artificial neural networks. The advantage of deep learning is to learn more useful features by building a machine learning model with multiple hidden layers and massive training data, so as to improve the accuracy of classification or prediction. It has a wide range of applications in image classification, voice recognition and text information screening in the Internet field, but it has not been involved in fishery, especially the feature recognition of koi seedlings. [0003] At present, existing domestic koi breeding ente...

AI Technical Summary

Problems solved by technology

[0004] The technical problem solved by the present invention is to provide a koi screening method and system device based on deep learning to solve the problem that koi seedlings are difficult to screen, improve screening accuracy, reduce manual use, and reduce the need for professional and technical personnel in breeding enterprises. dependence on

Images