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

Forestry pest recognition and detection method based on pooling vision Transformer

A detection method and forestry technology, applied in neural learning methods, character and pattern recognition, instruments, etc., can solve the problems of reduced recognition accuracy and stability, difficulty in segmentation, limited recognition ability, etc., and achieve high classification and recognition accuracy. , the effect of high accuracy and low image data requirements

Pending Publication Date: 2022-02-25
CHANGCHUN UNIV OF TECH
View PDF0 Cites 3 Cited by
  • Summary
  • Abstract
  • Description
  • Claims
  • Application Information

AI Technical Summary

Problems solved by technology

On the one hand, artificially designed image segmentation is difficult to obtain the feature expression closest to the natural attributes of the target, and the resulting recognition ability is limited
On the other hand, image preprocessing has an important impact on the recognition results, especially the segmentation of forestry pest eggs. For targets with a large degree of overlap, it is often difficult to segment or segment without affecting their respective feature expressions. , resulting in a decrease in the actual recognition accuracy and stability

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
  • Forestry pest recognition and detection method based on pooling vision Transformer
  • Forestry pest recognition and detection method based on pooling vision Transformer
  • Forestry pest recognition and detection method based on pooling vision Transformer

Examples

Experimental program
Comparison scheme
Effect test

example 1

[0027] Example 1 The present invention in the experimental results of forestry pest data set

[0028] The index that the present invention uses is, and following table is that experiment preliminary result compares:

[0029]

[0030] Time: The time it takes for the classifier to train every 20 Epoch, in seconds.

[0031]The method of the present invention is compared with the traditional convolutional neural network methods such as AlexNet, Vgg16, Resnet18, Densenet121 in the forestry pest data set, and also compared with the ViT (VisionTransformer) based on the attention mechanism, and the accuracy rate (Accuracy) is selected. , training time and parameters are used as evaluation indicators. The accuracy rate of the PiT method used in the present invention is 92.6%. The PiT model introduces a pooling operation on the basis of using the Transformer structure, fully considering the local feature information and global structural information of the image. Therefore, the PiT ...

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 forestry pest recognition and detection method based on pooling vision Transform. The method aims at discriminating and detecting different types of forestry pests in different periods. The method mainly comprises the steps: reasonably dividing data sets of different periods and different types into a training set and a verification set; expanding training set pictures by using a sample enhancement technology; proposing a forestry pest recognition feature extraction method based on a pre-trained PiT neural network model, adopting an improved transfer learning method to perform secondary training on an initialized model by using a new data set on the basis of the trained model to obtain a forestry pest recognition classifier; and performing fine adjustment on model parameters by using the test set. By adopting the method, the types and periods of the forestry pests can be quickly and accurately recognized, and the requirement on the acquired image data is low.

Description

technical field [0001] The invention belongs to the field of forestry pest identification, and in particular relates to a method for identifying and detecting forestry pests based on pooled vision Transformer. Background technique [0002] Forestry pests in farmland pose a great threat to the normal growth of crops and seriously affect the high and stable yield of crops. At present, the traditional extensive large-scale chemical insecticide has produced many negative effects, such as environmental pollution and threats to food safety. It can reduce the impact on the field ecological environment, reduce economic costs, and improve the efficiency of insect removal. Therefore, combined with the current development trend of automated agriculture, it is particularly important to automatically and efficiently identify forestry pests. [0003] In recent years, with the rapid development of computer vision technology, it provides a technical basis for the automatic recognition and...

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): G06V20/10G06V10/764G06V10/774G06V10/82G06K9/62G06N3/04G06N3/08
CPCG06N3/04G06N3/084G06F18/214G06F18/24
Inventor 王贵参杨承林王红梅
Owner CHANGCHUN UNIV OF TECH
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