Unlock instant, AI-driven research and patent intelligence for your innovation.

Method and device for joint identification of crop types and types of diseases and insect pests

A recognition method and crop technology, applied in character and pattern recognition, biological neural network models, instruments, etc., can solve the problems of time-consuming and labor-intensive human operation participation and limitations in collecting massive training data

Active Publication Date: 2021-10-01
宁夏智启连山科技有限公司
View PDF12 Cites 0 Cited by
  • Summary
  • Abstract
  • Description
  • Claims
  • Application Information

AI Technical Summary

Problems solved by technology

At present, many workers use image recognition technology to automatically identify crop diseases and insect pests, but these technologies are often limited by various factors such as light, angle, number of leaves, and background in the field.
At the same time, the development of deep learning in recent years has improved computer image recognition technology, but the process of collecting massive training data is time-consuming and laborious, and the data labeling process requires a large number of human operations.

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
  • Method and device for joint identification of crop types and types of diseases and insect pests
  • Method and device for joint identification of crop types and types of diseases and insect pests
  • Method and device for joint identification of crop types and types of diseases and insect pests

Examples

Experimental program
Comparison scheme
Effect test

no. 1 example

[0030] see figure 1 , the first embodiment of the present invention shows an identification system 10 . The recognition system 10 includes: a joint recognition device 100 , an image of a crop to be recognized 11 and a power supply device 12 .

[0031]The joint identification device 100 is connected to the power supply device 12, and the power supply device 12 provides the joint identification device 100 with the electric energy required for normal operation. The crop image 11 to be recognized is input into the joint recognition device 100 , and the joint recognition device 100 obtains the probability evaluation value of the corresponding crop type and pest type according to the crop image 11 to be recognized.

no. 2 example

[0033] see figure 2 , the embodiment of the present invention provides a method for joint identification of crop types and types of diseases and insect pests. The method steps include: step S100, step S200, step S300 and step S400.

[0034] Step S100: Using the preset deep full convolutional neural network to extract the features of the crop types from the image of the crops to be detected to obtain the feature map of the crop types and the feature of the pest types to obtain the feature map of the pest types.

[0035] Wherein, the step S100 includes:

[0036] 1. Carrying out a convolution operation on the crop image to obtain a crop feature map;

[0037] 2. Performing a large-area pooling operation on the crop feature map to obtain the crop type feature map and performing a small-area pooling operation to obtain the disease and insect pest type feature map.

[0038] The full convolutional neural network of the depth is used as an image feature extractor, and the deep conv...

no. 3 example

[0069] The second embodiment of the present invention provides a joint identification device for crop types and pest types, please refer to image 3 , Figure 4 , Figure 5 , Figure 6 and Figure 7, the joint identification device 100 includes: an image feature extraction module 110 , a first computing module 120 , a second computing module 130 and a third computing module 140 . The image feature extraction module 110 is used to use the preset depth full convolutional neural network to perform crop type feature extraction on the crop image to be detected to obtain a crop type feature map and perform disease and pest type feature extraction to obtain a disease and pest type feature map; the first The operation module 120 is used to operate the feature map of the disease and pest type through a preset multi-instance learning fusion model to obtain the first evaluation value of the disease and pest type, and obtain the crop according to the preset training set and the first e...

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 present invention provides a method and device for joint identification of crop types and pest types. In the method, the characteristic map of pest types is calculated through a preset multi-instance learning fusion model to obtain the first evaluation value of pest types. According to the preset The training set and the evaluation value of the first pest type obtain the Bayesian probability value of the crop category. Afterwards, the crop type feature map is passed through the preset fully connected neural network model to obtain the first crop type evaluation value, and the Bayesian probability value of the pest type is obtained according to the pre-stored training set and the first crop type evaluation value. Obtain a second pest type assessment value based on the first pest type assessment value and the Bayesian probability value of the pest type, and then obtain a second crop category assessment value based on the first crop category assessment value and the crop category Bayesian probability value. While realizing high-accuracy identification of crop types and types of diseases and insect pests, the demand for training data is greatly reduced.

Description

technical field [0001] The invention relates to the technical field of pest control, in particular to a method and device for joint identification of crop types and pest types. Background technique [0002] As one of the major agricultural disasters, crop diseases and insect pests seriously affect the effective progress of agricultural production activities. The identification of crop diseases and insect pests is of great significance to the control of crop diseases and insect pests. At present, many workers use image recognition technology to automatically identify crop diseases and insect pests, but these technologies are often limited by various factors such as light, angle, number of leaves, and background in the field. At the same time, the development of deep learning in recent years has improved computer image recognition technology, but the process of collecting massive training data is time-consuming and laborious, and the data labeling process requires a large num...

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 Patents(China)
IPC IPC(8): G06K9/62G06N3/04
CPCG06N3/045G06F18/24155G06F18/254G06F18/253
Inventor 张长水李易阳陆江
Owner 宁夏智启连山科技有限公司