Looking for breakthrough ideas for innovation challenges? Try Patsnap Eureka!

Bamboo forest pest identification method based on a convolutional neural network model

A technology of convolutional neural network and recognition method, which is applied in the field of recognition of bamboo forest pests based on convolutional neural network model, can solve the problems of insufficient data sample size, large fluctuation of model fitting degree, and low recognition accuracy, so as to avoid Large differences in body characteristics, automatic recognition and classification, and high recognition accuracy

Inactive Publication Date: 2019-04-02
ZHEJIANG FORESTRY UNIVERSITY
View PDF6 Cites 19 Cited by
  • Summary
  • Abstract
  • Description
  • Claims
  • Application Information

AI Technical Summary

Problems solved by technology

[0016] Purpose of the invention: In order to overcome the deficiencies in the prior art, the methods adopted in the prior art have insufficient data sample size, complex data preprocessing, insufficient feature extraction, large fluctuations in model fitting, and no specificity and complexity of pests. To optimize and adjust the model, all lead to low recognition accuracy

Method used

the structure of the environmentally friendly knitted fabric provided by the present invention; figure 2 Flow chart of the yarn wrapping machine for environmentally friendly knitted fabrics and storage devices; image 3 Is the parameter map of the yarn covering machine
View more

Image

Smart Image Click on the blue labels to locate them in the text.
Viewing Examples
Smart Image
  • Bamboo forest pest identification method based on a convolutional neural network model
  • Bamboo forest pest identification method based on a convolutional neural network model
  • Bamboo forest pest identification method based on a convolutional neural network model

Examples

Experimental program
Comparison scheme
Effect test

Embodiment 1

[0066] A kind of recognition method of bamboo forest pests based on convolutional neural network model of the present embodiment, refer to figure 2 , including the following steps:

[0067] Collect image data of bamboo forest pest samples;

[0068] Processing the sample image data of bamboo forest pests to obtain a processed image data set of bamboo forest pest samples;

[0069] Use the VGG convolutional neural network model and optimize the VGG convolutional neural network model:

[0070] VGG convolutional neural network model reference figure 1 , mainly composed of input layer, convolutional layer, pooling layer, fully connected layer, and Softmax layer. VGGNet can be divided into 5 layer groups, including Conv1 convolutional layer group 1, Conv2 convolutional layer group 2, Conv3 convolutional layer group 3, Conv4 convolutional layer group 4 and Conv5 convolutional layer group 5, fully connected layer including FC6 Fully connected layer 6 and FC7 fully connected layer ...

Embodiment 2

[0077] A kind of recognition method of bamboo forest pests based on the convolutional neural network model of this embodiment, based on embodiment one, collects the image data of different growth stages of bamboo forest pests, for example, collects the image data of pests from eggs to larvae to adult stages, for training The model avoids the error in model detection caused by the large difference in the body characteristics of pests in different growth stages. Of course, the types of pests are different and the growth stages of pests are different. In addition, the color of the bamboo forest pest sample image is retained according to the situation, because the pest not only has differences in shape and texture, but color is also an important factor affecting the recognition accuracy. Therefore, in the data processing process, the input image data is not converted into a grayscale image. , but retains the values ​​of the RGB three color channels of the image.

Embodiment 3

[0079] A kind of bamboo forest pest recognition method based on convolutional neural network model of the present embodiment, based on embodiment two, bamboo forest pest sample image data is processed, obtains the bamboo forest pest sample image data set after processing, comprises the following steps:

[0080] Handle erroneous and repeated data in the image data of bamboo forest pest samples;

[0081] The image data of bamboo forest pest samples are expanded by means of data augmentation.

[0082]Expand the sample image data of bamboo forest pests by means of data enhancement: After screening, the number of samples of each insect is different. In order to improve the classification accuracy of the network, make the performance of the network better, and prevent problems such as overfitting, this embodiment Adopt the mode of data enhancement to expand the above-mentioned bamboo forest pest sample image data amount, by operating the above-mentioned bamboo forest pest sample ima...

the structure of the environmentally friendly knitted fabric provided by the present invention; figure 2 Flow chart of the yarn wrapping machine for environmentally friendly knitted fabrics and storage devices; image 3 Is the parameter map of the yarn covering machine
Login to View More

PUM

No PUM Login to View More

Abstract

The invention discloses a bamboo forest pest recognition method based on a convolutional neural network model, and belongs to the technical field of agricultural and forestry pest recognition and classification. The method comprises the steps of collecting the bamboo forest pest sample image data; processing the pest sample image data of the bamboo forest to obtain a processed pest sample image data set of the bamboo forest; adopting a VGG convolutional neural network model, and optimizing the VGG convolutional neural network model; carrying out target pest identification and classification byutilizing the optimized VGG convolutional neural network model; developing a set of pest identifying and classifying method suitable for bamboo forests based on a VGGNet network structure, so that rapid identification and classification of common bamboo forest pests can be achieved, the automatic identification and classification can be achieved, and the identification accuracy is high.

Description

technical field [0001] The invention belongs to the technical field of identification and classification of agricultural and forestry pests, in particular to a bamboo forest pest identification method based on a convolutional neural network model. Background technique [0002] Our country is a large agricultural country. During the cultivation of crops, different kinds of pests are encountered every year, which makes the yield and quality of crops decline to varying degrees. When the disaster is severe, it may even lead to a large area of ​​crop failure. Accurate and effective classification, identification and identification of insects is an important prerequisite for timely pest control and avoiding huge economic losses of crops. Insects are the most diverse animals in the natural environment, and it is very difficult to identify them due to their varied shapes and rich textures. Traditional insect taxonomy and identification mainly rely on insect experts or insect taxono...

Claims

the structure of the environmentally friendly knitted fabric provided by the present invention; figure 2 Flow chart of the yarn wrapping machine for environmentally friendly knitted fabrics and storage devices; image 3 Is the parameter map of the yarn covering machine
Login to View More

Application Information

Patent Timeline
no application Login to View More
Patent Type & Authority Applications(China)
IPC IPC(8): G06K9/00G06N3/04G06N3/08
CPCG06N3/084G06V20/10G06N3/045
Inventor 冯海林任丽锦方益明杨垠晖刘兴泉周国鑫
Owner ZHEJIANG FORESTRY UNIVERSITY
Who we serve
  • R&D Engineer
  • R&D Manager
  • IP Professional
Why Patsnap Eureka
  • Industry Leading Data Capabilities
  • Powerful AI technology
  • Patent DNA Extraction
Social media
Patsnap Eureka Blog
Learn More
PatSnap group products

Browse by: Latest US Patents, China's latest patents, Technical Efficacy Thesaurus, Application Domain, Technology Topic, Popular Technical Reports.

© 2024 PatSnap. All rights reserved.Legal|Privacy policy|Modern Slavery Act Transparency Statement|Sitemap|About US| Contact US: help@patsnap.com