Seed sorting method for sunflower crop on basis of deep convolutional neural network

Abstract

The invention provides a seed sorting method for a sunflower crop on the basis of a deep convolutional neural network. The seed sorting method comprises the steps of: firstly, collecting original RGBsunflower seed image data, carrying out marking, randomly dividing the original RGB sunflower seed image data into a training sample set and a verification sample set, and carrying out amplification and standardization on the training sample set and a verification sample set; secondly, constructing a deep convolutional neural network by utilizing standard convolution, a residual module containingan attention mechanism module, a pooling layer and a classifier, training the training sample set, and optimizing network parameters in the deep convolutional neural network by combining a random gradient descent algorithm so as to obtain a deep convolutional neural network model; and finally, verifying the deep convolutional neural network model by using the verification sample set, and testing the recognition capability of the model. According to the invention, the memory required for training the model is reduced; features with high robustness can be automatically learned and extracted froma color sample image of sunflower seeds; meanwhile, the method has high recognition rate.

Description

technical field [0001] The invention relates to the technical field of automatic sorting of crop seeds, in particular to a sunflower crop seed sorting method based on a deep convolutional neural network. Background technique [0002] Automatic sorting of crop seeds is an important task in automated or semi-automated industrial production. The quality of crop seeds will have an important impact on food quality issues. However, during the harvesting and storage of crops, a large number of sundries will be mixed in, such as leaves, stems, moldy crop seeds, etc., which will affect the quality of subsequent processed products. . Therefore, correct identification and removal of undesirable seeds, stones, leaves and other impurities is the key to ensuring product quality. However, due to the random distribution of sunflower crop seed images, the difference between various bad sunflower seeds and other debris mixed with normal sunflower seeds is often not obvious, which greatly in...

AI Technical Summary

Problems solved by technology

Although the method based on convolutional neural network is robust and has high recognition accuracy, the storage space of these models trained by convolutional neural network often requires hundreds of megabytes and a long training time. The production is limited by the storage space of the hardware device

Images