Fruit tree pest image recognition method
A technology of image recognition and pests and diseases, which is applied in the field of image recognition of fruit tree diseases and insect pests, can solve the problems of difficult large-scale promotion and long detection time, and achieve the effect of flexibility, good classification effect and high recognition accuracy
Pending Publication Date: 2022-07-05
YULIN NORMAL UNIVERSITY
View PDF0 Cites 2 Cited by
- Summary
- Abstract
- Description
- Claims
- Application Information
AI Technical Summary
Problems solved by technology
This type of method has a high accuracy rate for identifying citrus diseases and insect pests, but because the detection requires relevant detection equipment and professional technicians to complete, the detection time is long and it is difficult to promote it on a large scale
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 moreImage
Smart Image Click on the blue labels to locate them in the text.
Smart ImageViewing Examples
Examples
Experimental program
Comparison scheme
Effect test
Embodiment Construction
[0038] The present invention will be further described below with reference to the specific embodiments in the accompanying drawings.
[0039] VGG19 is a classic convolutional neural network developed by researchers at Oxford University and Google. It adopts a cascaded network structure and consists of 16 convolutional layers and 3 fully connected layers; each convolutional layer uses a 3×3 convolution kernel, and there are a total of 5 maximum pooling layers between the convolutional layers; The last layer is the Softmax classifier. Since the VGG19 model uses a fixed 3 × 3 convolution kernel, its receptive field is fixed, and it is easy to ignore some subtle features, which in turn results in insufficient fine-grained feature extraction. Therefore, when using this model to identify and classify the early citrus Huanglongbing, nematode disease images and healthy leaves with high similarity, the recognition rate is low. In addition, the VGG19 model uses 3 fully-connected laye...
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
Login to View More Abstract
The invention discloses a fruit tree disease and insect pest image recognition method, relates to the technical field of agricultural planting, and solves the technical problems of poor accuracy and long detection time of an existing citrus disease and insect pest recognition and diagnosis method. The method comprises the following steps: acquiring a pest image sample, and marking; performing enhancement processing on the pest image sample by using an image enhancement method to obtain an enhanced data set; pre-training the VGG19 model by using an ImageNet data set to obtain a pre-training weight parameter; a VGG19-INC model is constructed by reserving the first four convolution layers and pooling layers of a VGG19 model, replacing the fifth convolution layer of the VGG19 model with one batch standardized convolution layer and two Inception modules, replacing a full connection layer of the VGG19 model with one global pooling layer and finally constructing by using a 1 * 4 Softmax layer; using the enhanced data set to train a VGG19-INC model; and applying the trained VGG19-INC model to classification prediction of unknown pest and disease images, and finally obtaining a pest and disease image identification result.
Description
technical field [0001] The invention relates to the technical field of agricultural planting, and more particularly, to a method for identifying images of fruit tree diseases and insect pests. Background technique [0002] The diseases and insect pests of citrus can be divided into invasive and non-infectious diseases, as well as three types of insect pests. Especially, Huanglongbing, leaf miner and nematode have the greatest impact on citrus planting. The end of the area. Therefore, the rapid and accurate identification and diagnosis of citrus diseases and insect pests can provide a basis for later treatment, which has become an urgent problem to be solved in the current citrus planting industry. At present, the identification and diagnosis methods of citrus diseases and insect pests mainly include visual diagnosis method and pathological characteristic detection method. The visual diagnosis method relies on the on-site visual inspection of farmers or agricultural technic...
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
Login to View More Patent Type & Authority Applications(China)
IPC IPC(8): G06V20/10G06T7/00G06V10/764G06V10/774G06V10/82G06N3/04G06N3/08
CPCG06T7/0012G06N3/08G06T2207/20084G06T2207/20081G06T2207/30004G06N3/047G06N3/045G06F18/2431G06F18/2415G06F18/214
Inventor 闭吕庆黄平
Owner YULIN NORMAL UNIVERSITY