CN115294555A - Plant disease intelligent diagnosis method and system based on neural network - Google Patents
Plant disease intelligent diagnosis method and system based on neural network Download PDFInfo
- Publication number
- CN115294555A CN115294555A CN202211182670.0A CN202211182670A CN115294555A CN 115294555 A CN115294555 A CN 115294555A CN 202211182670 A CN202211182670 A CN 202211182670A CN 115294555 A CN115294555 A CN 115294555A
- Authority
- CN
- China
- Prior art keywords
- image
- neural network
- color difference
- plant
- color
- Prior art date
- Legal status (The legal status is an assumption and is not a legal conclusion. Google has not performed a legal analysis and makes no representation as to the accuracy of the status listed.)
- Granted
Links
Images
Classifications
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06V—IMAGE OR VIDEO RECOGNITION OR UNDERSTANDING
- G06V20/00—Scenes; Scene-specific elements
- G06V20/60—Type of objects
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06N—COMPUTING ARRANGEMENTS BASED ON SPECIFIC COMPUTATIONAL MODELS
- G06N3/00—Computing arrangements based on biological models
- G06N3/02—Neural networks
- G06N3/08—Learning methods
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06V—IMAGE OR VIDEO RECOGNITION OR UNDERSTANDING
- G06V10/00—Arrangements for image or video recognition or understanding
- G06V10/40—Extraction of image or video features
- G06V10/56—Extraction of image or video features relating to colour
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06V—IMAGE OR VIDEO RECOGNITION OR UNDERSTANDING
- G06V10/00—Arrangements for image or video recognition or understanding
- G06V10/70—Arrangements for image or video recognition or understanding using pattern recognition or machine learning
- G06V10/764—Arrangements for image or video recognition or understanding using pattern recognition or machine learning using classification, e.g. of video objects
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06V—IMAGE OR VIDEO RECOGNITION OR UNDERSTANDING
- G06V10/00—Arrangements for image or video recognition or understanding
- G06V10/70—Arrangements for image or video recognition or understanding using pattern recognition or machine learning
- G06V10/82—Arrangements for image or video recognition or understanding using pattern recognition or machine learning using neural networks
Landscapes
- Engineering & Computer Science (AREA)
- Theoretical Computer Science (AREA)
- Physics & Mathematics (AREA)
- General Physics & Mathematics (AREA)
- Evolutionary Computation (AREA)
- Multimedia (AREA)
- Health & Medical Sciences (AREA)
- Computing Systems (AREA)
- Artificial Intelligence (AREA)
- General Health & Medical Sciences (AREA)
- Software Systems (AREA)
- Computer Vision & Pattern Recognition (AREA)
- Databases & Information Systems (AREA)
- Medical Informatics (AREA)
- Biomedical Technology (AREA)
- Life Sciences & Earth Sciences (AREA)
- Biophysics (AREA)
- Computational Linguistics (AREA)
- Data Mining & Analysis (AREA)
- Molecular Biology (AREA)
- General Engineering & Computer Science (AREA)
- Mathematical Physics (AREA)
- Image Analysis (AREA)
Abstract
The invention relates to a plant disease intelligent diagnosis method and system based on a neural network, which are used for acquiring plant leaf images, wherein the plant leaf images comprise plant leaf images to be identified and healthy leaf images; constructing an antagonistic neural network, and training the antagonistic neural network by using the healthy blade image to obtain a trained antagonistic neural network; inputting the plant leaf image to be identified into a trained antagonistic neural network to obtain a reconstructed color image; carrying out color difference calculation on the reconstructed color image and the plant leaf image to be identified to obtain a reconstructed color difference image; inputting the reconstructed color difference image into the constructed convolutional neural network, and outputting a classification result; and obtaining a plant disease diagnosis result according to the classification result. Namely, the scheme of the invention can utilize the reconstructed color difference image to diagnose the plant diseases by obtaining the reconstructed color difference image, can visualize the disease area on the pixel level and reduce the calculated amount.
Description
Technical Field
The invention relates to the technical field of intelligent agricultural detection, in particular to a plant disease intelligent diagnosis method and system based on a neural network.
Background
In various natural disasters of China, crop diseases and insect pests occupy a very important position, so that the plant diseases and insect pests need to be predicted and monitored in time to prevent serious disasters.
The traditional machine learning method generally comprises the steps of lesion extraction, edge feature extraction and the like, and finally, classification is carried out by using a support vector machine. However, the traditional machine learning method has many and complex preprocessing steps for images, can only be applied to a small amount of pest detection of individual plants, and has poor mobility and low accuracy.
Disclosure of Invention
The invention aims to provide a plant disease intelligent diagnosis method and system based on a neural network, which are used for solving the problems of poor universality, complex identification method and unstable identification rate among different plants in the existing detection technology.
In order to achieve the purpose, the technical scheme of the plant disease intelligent diagnosis method based on the neural network comprises the following steps:
acquiring a plant leaf image, wherein the plant leaf image comprises a plant leaf image to be identified and a healthy leaf image;
constructing an antagonistic neural network, and training the antagonistic neural network by using the healthy leaf images to obtain the trained antagonistic neural network; inputting the plant leaf image to be identified into a trained antagonistic neural network to obtain a reconstructed color image; carrying out color difference calculation on the reconstructed color image and the plant leaf image to be identified to obtain a reconstructed color difference image;
inputting the reconstructed color difference image into the constructed convolutional neural network, and outputting a classification result; and obtaining a plant disease diagnosis result according to the classification result.
Further, the method also comprises the step of carrying out background removal processing on the plant leaf image.
Further, the antagonistic neural network comprises two neural networks, one being a generator and one being a discriminator;
wherein the loss function of the generator is:
wherein, the first and the second end of the pipe are connected with each other,L SSIM is composed ofSSIMA loss function;L Color 
a color loss function;
whereinnWhich represents the number of pixels of the image,w i is shown asiThe weight of each pixel is determined by the weight of the pixel,P i for reconstructing the first in color imageiThe grey values of the R, G, B color space of the individual pixels,T i for the image of the plant leaf to be identifiediThe gray values of the R, G, B color space of the individual pixels.
Further, the weight isw i The acquisition method comprises the following steps:
wherein the content of the first and second substances,
to remove any pixel value in the background healthy leaf image,C1 is the value of R, G and B channels of the main color of the blade,athe value range is (0, 1), and b has a value of 3.
Further, the reconstructed color difference image is:
wherein the content of the first and second substances,P_RGBin order to reconstruct a color image,T_RGBis the plant leaf image to be identified.
Further, the convolutional neural network comprises a feature extraction coder and a fully connected network; the feature extraction encoder comprises a plurality of convolution layers and an attention model;
the attention model is:
whereinDiffCColor difference values of pixels of the reconstructed color difference image; cmax is the maximum color difference value in the reconstructed color difference image.
The invention also provides a plant disease intelligent diagnosis system based on the neural network, which comprises a processor and a memory, wherein the processor is used for executing the technical scheme of the plant disease intelligent diagnosis system based on the neural network stored in the memory.
Compared with the prior art, the invention has the beneficial effects that:
according to the method, all leaf images do not need to be marked, and the disease area detection is carried out on the pixel level through color reconstruction, namely through the obtained reconstructed color difference image; meanwhile, by increasing the weight of the color reconstruction loss, the gradient of the reconstruction error of the pixel close to the main color of the blade in the blade image can be increased, the attention of the network to the reconstruction of the main color of the blade is improved, and the color of the blade image is more completely reconstructed.
The method of the invention also adds attention to the characteristic diagram of the convolutional neural network based on the reconstructed chromatic aberration image, so that the network training is more targeted and has stronger robustness, and the problem of poor classification precision caused by complicated and large-change positions of the leaves suffering from diseases is avoided.
Drawings
In order to more clearly illustrate the embodiments of the present specification or the technical solutions in the prior art, the drawings needed to be used in the embodiments will be briefly described below, and it is obvious that the drawings in the following description are only some embodiments of the present specification, and it is obvious for those skilled in the art that other drawings can be obtained according to the drawings without creative efforts.
FIG. 1 is a flowchart of a method of an embodiment of the plant disease intelligent diagnosis method based on neural network of the present invention;
FIG. 2 is a schematic diagram of an image being preprocessed to remove background;
fig. 3 is a reconstructed color difference thermodynamic diagram of an image.
Detailed Description
In the following, the technical solutions in the embodiments of the present invention will be clearly and completely described with reference to the drawings in the embodiments of the present invention, and it is obvious that the described embodiments are only a part of the embodiments of the present invention, rather than all of the embodiments, and all other embodiments obtained by a person skilled in the art based on the embodiments of the present invention belong to the protection scope of the embodiments of the present invention.
The invention provides a specific implementation mode of a plant disease intelligent diagnosis method based on a neural network.
Specifically, as shown in fig. 1, the intelligent plant disease diagnosis method based on the neural network comprises the following steps:
step 1, obtaining plant leaf images, wherein the plant leaf images comprise plant leaf images to be identified and healthy leaf images;
in the embodiment, the method further comprises the steps of carrying out image preprocessing on the obtained plant leaf image, wherein the image preprocessing comprises image enhancement, image denoising and the like; since image enhancement and image denoising are the prior art, they are not described in detail herein.
Further, background removal is performed on the preprocessed image, a watershed algorithm is preferably adopted in this embodiment, as shown in fig. 2, a bright pixel is regarded as a high altitude, a dark pixel is regarded as a low altitude to find a watershed in the image, so as to obtain a binary image, and then an adjacent region is segmented by using an eight neighborhood principle; finally, removing the small pixel groups through morphological processing to improve the mask; and finally, multiplying the processed binary image with the original image, reserving the leaf area, and obtaining a visible light image of the leaf area.
The plant leaf images obtained in the embodiment comprise plant leaf images to be identified and healthy leaf images; the plant leaf images to be identified are a plurality of plant leaf images shot by a camera; the healthy leaf images can be plant leaf images acquired in advance, and can also be plant leaf data stored in history directly.
Step 2, constructing an antagonistic neural network, and training the antagonistic neural network by using the healthy leaf images to obtain the trained antagonistic neural network; inputting the plant leaf image to be identified into a trained antagonistic neural network to obtain a reconstructed color image;
in the embodiment, the healthy leaf images are converted into gray images, and training of a resistance neural network is performed; the training process of the anti-neural network is as follows:
constructing an antagonistic neural network; the anti-neural network comprises a generator and a discriminator.
The generator is in encoder-decoder structure and can adoptUnetAnd (3) waiting for the network model, inputting the gray image of the leaf region into the encoder, outputting the gray image as a characteristic diagram, inputting the characteristic diagram into the decoder, and outputting the characteristic diagram as a visible light image of the leaf region, namely a reconstructed color image after fitting and upsampling. The generator reconstructs the color which accords with the real distribution of the training data under the guidance of the discriminator by learning the characteristics of the training set data, thereby generating the similar data with the characteristics of the training set. And the discriminator is responsible for distinguishing whether the input data is real or false data generated by the generator and feeding back the data to the generator.
The two neural networks are alternately trained, and the capacity is synchronously improved until the data generated by the generated network can be falsified and truthful and reach a certain balance with the capacity of the discrimination network.
The loss function of the generator is:
wherein the content of the first and second substances,L SSIM is composed ofSSIM(structurally similar) loss functions;L Color 
a color loss function.
For theL SSIM Function: the indexes of brightness (luminance), contrast (contrast) and structure (structure) are considered; due to the fact thatL SSIM The function formula is well known and will not be described in detail here.
For theL Color The function is:
whereinnWhich represents the number of pixels of the image,w i denotes the firstiThe weight of each of the pixels is determined,
for reconstructing the first in a color image
The grey values of the R, G, B color space of the individual pixels,T i for the image of the plant leaf to be recognizediThe values of the R, G, B color space of the individual pixels.
Wherein the weight isw i The acquisition method comprises the following steps:
1) Converting the healthy plant leaf images without the background into vectors, and clustering the obtained vectors to obtain two categories and clustering centers corresponding to the categoriesThe coordinates of one cluster center which is the main color of the leaf are recorded asC1 (i.e., the values of the three channels R, G, B); the other cluster center coordinate is the background color in the leaf region image, i.e., black.
In this example, use is made ofK-meansClustering by a clustering method; assume an image size ofM*N*3The conversion vector is [ alpha ], [ alpha ]M*N,3]Since the image is a visible image of the leaf area with background removed, and healthy leaves are usually one dominant color, the class of the cluster is K =2.
The weight is obtained, and the leaves comprise red leaves, yellow leaves and blue leaves besides green leaves; most leaves of plants are green in color because the leaves contain chloroplast organelles. The chloroplast contains pigments such as chlorophyll, lutein, carotene, etc. Usually, the chlorophyll content of the leaves is an absolute dominance, and the leaves are green, so that the colors are clustered, and the main color is taken as the main color of the leaves.
2) Calculating a weight for each pixel of the leaf regionw i Comprises the following steps:
wherein the content of the first and second substances,Pixelto remove any pixel value in the background healthy leaf image,athe value range is (0, 1),athe value is preferably 0.5, the value of b is preferably 3, i.e. the weight becomes 4 when the color difference is 1.
In this embodiment, in order to avoid that the training cannot be normally converged due to too high weight, the empirical value of c is 5; the minimum weight is 1.
As described aboveL Color In a functioncolorDiffenceFor the evaluation function of the chromatic aberration, the existing one is preferably usedCECIE2000A color difference formula; it should be noted that, by using the above color difference formula, the pixel values need to be converted into the color difference values in advanceLabA spatially corresponding value.
In the invention, by increasing the weight, the gradient of the reconstruction error of the pixel close to the main color of the blade in the blade image can be increased, and the attention of the network to the reconstruction of the main color of the blade is improved.
In this embodiment, the discriminator may use a common network model, such asResNet、DenseNetAnd the characteristic extraction is carried out on the image, and the input is carried out on a blade area gray level image and a blade area visible light imageConcatAnd (4) outputting a classification result, wherein the classification result is true or false, namely whether the image is a true image or not, and the image generated by the generator is a false image. The loss function adopts a cross entropy loss function.
There are many kinds of anti-neural networks, which can adopt the existing onesPix2pixNetwork models, etc.
The generator in the above-mentioned antagonistic neural network is generally a hidden structure which can capture normal data well only by training the normal data. After training, the generator can reconstruct normal data well, but not the abnormal samples correctly. Thus, by comparing the reconstruction error between the test sample and the reconstructed sample, anomalies can be detected; i.e. the reconstruction error of normal samples is typically smaller than the reconstruction error of abnormal samples. Therefore, the inference of the network generator is carried out on each visible light image of the leaf area, and a reconstructed color image is obtained.
Step 3, carrying out color difference calculation on the reconstructed color image and the plant leaf image to be identified to obtain a reconstructed color difference image;
the reconstructed color difference image is as follows:
wherein, the first and the second end of the pipe are connected with each other,P_RGBin order to reconstruct a color image,T_RGBis the plant leaf image to be identified.
In this embodiment, the existing ones are adoptedCECIE2000Calculating a color difference value by a color difference formula, wherein when the calculated color difference value is larger, the color difference is larger; then, color mapping is performed on the reconstructed color difference image to obtain a reconstructed color difference thermodynamic diagram, as shown in fig. 3And providing visualization for disease diagnosis of plant leaves.
Step 4, inputting the reconstructed color difference image into the constructed convolutional neural network, and outputting a classification result; and obtaining a plant disease diagnosis result according to the classification result.
In this embodiment, a convolutional neural network is constructed, where the network includes a feature extraction encoder and a fully-connected network, and the feature extraction encoder includes a plurality of convolutional layers and an attention model.
Inputting a reconstructed color difference image into a convolutional neural network, performing Feature extraction on the input reconstructed color difference image through a Feature extraction encoder, outputting the reconstructed color difference image into Feature maps under multiple resolutions, then performing attention addition on Feature maps through an attention model, performing downsampling on the image to obtain attention images of different scales, and multiplying the attention images with the Feature maps of corresponding scales respectively to obtain an attention Feature map; inputting the attention feature map into a fully-connected network, and outputting a classification result; and performing argmax operation on the classification result to obtain a plant disease diagnosis result.
The fully connected network in this embodiment serves to map features to the sample label space.
In the embodiment, the number of the attention feature maps is 4, namely 4 feature maps with different scales are obtained; the classification result output by the embodiment is a probability.
The plant diagnosis results in this example were rust, brown spot, powdery mildew, anthracnose, black spot, leaf spot, blight, angular leaf spot, puncture disease, and blight, all of which were 11 types.
The attention model in this example is:
whereinDiffCColor difference values of pixels of the reconstructed color difference image; cmax is the maximum color difference value in the reconstructed color difference image. The 1 in the above formula is to ensure that the characteristic value is not reduced and avoid the phenomenon that the gradient disappears in the training process。
In this embodiment, the training method of the constructed convolutional neural network is as follows:
acquiring a plant leaf image with diseases, and marking the plant leaf image with the diseases to obtain a label data set;
inputting the label data set into the constructed convolutional neural network for training;
wherein the loss function of the convolutional neural network adopts a cross entropy loss function.
The labels of the plant diseases in this embodiment need to be set for scenes, for example, the diseases are defined as rust disease, brown spot, powdery mildew, anthracnose, black spot, leaf spot, blight, angular leaf spot, perforation disease and dry branch disease, which are normal, and are totally 11 types, and 0-10 arabic numerals are respectively used as the corresponding labels of the diseases; wherein the tag data needs to be one-hot encoded before being input into the network for training.
In the embodiment, as the larger the chromatic aberration value in the reconstructed chromatic aberration image is, the larger the abnormality is, and the complex and large variation of the position of the blade generally suffering from the disease is, the attention of the characteristic diagram of the neural network can be added based on the reconstructed chromatic aberration image, so that the network training is more targeted, and the robustness is stronger.
There are many kinds of convolutional neural network structures, and the invention adoptsResNetOr the lightweight network pre-training model can improve the convergence efficiency and the classification precision.
Based on the same inventive concept as the method, the invention also provides a plant disease intelligent diagnosis system based on the neural network, which comprises a processor and a memory, wherein the processor is used for executing the program of the plant disease intelligent diagnosis method embodiment based on the neural network stored in the memory; since the embodiment of the plant disease intelligent diagnosis method based on the neural network has been described in the above embodiment, redundant description is not repeated here.
As will be appreciated by one skilled in the art, embodiments of the present invention may be provided as a method, system, or computer program product. Accordingly, the present invention may take the form of an entirely hardware embodiment, an entirely software embodiment or an embodiment combining software and hardware aspects. Furthermore, the present invention may take the form of a computer program product embodied on one or more computer-usable storage media (including, but not limited to, disk storage, CD-ROM, optical storage, and the like) having computer-usable program code embodied therein.
The present invention is described with reference to flowchart illustrations and/or block diagrams of methods, apparatus (modules, systems) and computer program products according to embodiments of the invention. It will be understood that each flow and/or block of the flowchart illustrations and/or block diagrams, and combinations of flows and/or blocks in the flowchart illustrations and/or block diagrams, can be implemented by computer program instructions. These computer program instructions may be provided to a processor of a general purpose computer, special purpose computer, embedded computer, or other programmable data processing apparatus to produce a machine, such that the instructions, which execute via the processor of the computer or other programmable data processing apparatus, create means for implementing the functions specified in the flowchart flow or flows and/or block diagram block or blocks.
These computer program instructions may also be stored in a computer-readable memory that can direct a computer or other programmable data processing apparatus to function in a particular manner, such that the instructions stored in the computer-readable memory produce an article of manufacture including instruction means which implement the function specified in the flowchart flow or flows and/or block diagram block or blocks.
These computer program instructions may also be loaded onto a computer or other programmable data processing apparatus to cause a series of operational steps to be performed on the computer or other programmable apparatus to produce a computer implemented process such that the instructions which execute on the computer or other programmable apparatus provide steps for implementing the functions specified in the flowchart flow or flows and/or block diagram block or blocks.
The embodiments in the present specification are described in a progressive manner, and the same and similar parts among the embodiments are referred to each other, and each embodiment focuses on the differences from the other embodiments.
The above description is only for the purpose of illustrating the preferred embodiments of the present invention and is not to be construed as limiting the invention, and any modifications, equivalents, improvements and the like that fall within the spirit and principle of the present invention are intended to be included therein.
Claims (7)
1. The intelligent plant disease diagnosis method based on the neural network is characterized by comprising the following steps of:
acquiring a plant leaf image, wherein the plant leaf image comprises a plant leaf image to be identified and a healthy leaf image;
constructing an antagonistic neural network, and training the antagonistic neural network by using the healthy leaf images to obtain the trained antagonistic neural network; inputting the plant leaf image to be identified into a trained antagonistic neural network to obtain a reconstructed color image; carrying out color difference calculation on the reconstructed color image and the plant leaf image to be identified to obtain a reconstructed color difference image;
inputting the reconstructed color difference image into the constructed convolutional neural network, and outputting a classification result; and obtaining a plant disease diagnosis result according to the classification result.
2. The intelligent plant disease diagnosis method based on the neural network according to claim 1, further comprising a step of performing background removal processing on the plant leaf image.
3. The intelligent plant disease diagnosing method based on neural network as claimed in claim 1, wherein the antagonistic neural network includes a generator and a discriminator;
wherein the loss function of the generator is:
wherein the content of the first and second substances,L SSIM is composed ofSSIMA loss function;L Color 
a color loss function;
whereinnWhich represents the number of pixels of the image,w i is shown asiThe weight of each pixel is determined by the weight of the pixel,P i for reconstructing the first in color imageiThe grey values of the R, G, B color space of the individual pixels,T i for the image of the plant leaf to be identifiediThe gray values of the R, G, B color space of the individual pixels.
4. The intelligent plant disease diagnosis method based on neural network as claimed in claim 3, wherein the weight is set as weightw i The acquisition method comprises the following steps:
wherein the content of the first and second substances,Pixelto remove any pixel value in the background healthy leaf image,C1 is a coordinate point corresponding to the values of the R, G and B channels of the main color of the blade,athe value range is (0, 1), and b has a value of 3.
5. The intelligent plant disease diagnosis method based on neural network as claimed in claim 1, wherein the reconstructed color difference image is:
wherein, the first and the second end of the pipe are connected with each other,P_RGBin order to reconstruct a color image,T_RGBis the plant leaf image to be identified.
6. The intelligent plant disease diagnosis method based on the neural network as claimed in claim 1, wherein the convolutional neural network comprises a feature extraction encoder and a fully connected network; the feature extraction encoder comprises a plurality of convolution layers and an attention model;
the attention model is:
whereinDiffCColor difference values of pixels of the reconstructed color difference image; cmax is the maximum color difference value in the reconstructed color difference image.
7. A plant disease intelligent diagnosis system based on a neural network, which comprises a processor and a memory, and is characterized in that the processor is used for executing instructions stored by the memory for implementing the plant disease intelligent diagnosis method based on the neural network according to any one of claims 1-6.
Priority Applications (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN202211182670.0A CN115294555B (en) | 2022-09-27 | 2022-09-27 | Plant disease intelligent diagnosis method and system based on neural network |
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN202211182670.0A CN115294555B (en) | 2022-09-27 | 2022-09-27 | Plant disease intelligent diagnosis method and system based on neural network |
Publications (2)
| Publication Number | Publication Date |
|---|---|
| CN115294555A true CN115294555A (en) | 2022-11-04 |
| CN115294555B CN115294555B (en) | 2022-12-20 |
Family
ID=83835005
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CN202211182670.0A Active CN115294555B (en) | 2022-09-27 | 2022-09-27 | Plant disease intelligent diagnosis method and system based on neural network |
Country Status (1)
| Country | Link |
|---|---|
| CN (1) | CN115294555B (en) |
Cited By (1)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN116782041A (en) * | 2023-05-29 | 2023-09-19 | 武汉工程大学 | Image quality improvement method and system based on liquid crystal microlens array |
Citations (5)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN110210434A (en) * | 2019-06-10 | 2019-09-06 | 四川大学 | Pest and disease damage recognition methods and device |
| CN111724372A (en) * | 2020-06-19 | 2020-09-29 | 深圳新视智科技术有限公司 | Method, terminal and storage medium for detecting cloth defects based on antagonistic neural network |
| CN112183635A (en) * | 2020-09-29 | 2021-01-05 | 南京农业大学 | Method for realizing segmentation and identification of plant leaf lesions by multi-scale deconvolution network |
| CN112699941A (en) * | 2020-12-31 | 2021-04-23 | 浙江科技学院 | Plant disease severity image classification method and device, computer equipment and storage medium |
| CN114548265A (en) * | 2022-02-21 | 2022-05-27 | 安徽农业大学 | Crop leaf disease image generation model training method, crop leaf disease identification method, electronic device and storage medium |
-
2022
- 2022-09-27 CN CN202211182670.0A patent/CN115294555B/en active Active
Patent Citations (5)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN110210434A (en) * | 2019-06-10 | 2019-09-06 | 四川大学 | Pest and disease damage recognition methods and device |
| CN111724372A (en) * | 2020-06-19 | 2020-09-29 | 深圳新视智科技术有限公司 | Method, terminal and storage medium for detecting cloth defects based on antagonistic neural network |
| CN112183635A (en) * | 2020-09-29 | 2021-01-05 | 南京农业大学 | Method for realizing segmentation and identification of plant leaf lesions by multi-scale deconvolution network |
| CN112699941A (en) * | 2020-12-31 | 2021-04-23 | 浙江科技学院 | Plant disease severity image classification method and device, computer equipment and storage medium |
| CN114548265A (en) * | 2022-02-21 | 2022-05-27 | 安徽农业大学 | Crop leaf disease image generation model training method, crop leaf disease identification method, electronic device and storage medium |
Cited By (2)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN116782041A (en) * | 2023-05-29 | 2023-09-19 | 武汉工程大学 | Image quality improvement method and system based on liquid crystal microlens array |
| CN116782041B (en) * | 2023-05-29 | 2024-01-30 | 武汉工程大学 | Image quality improvement method and system based on liquid crystal microlens array |
Also Published As
| Publication number | Publication date |
|---|---|
| CN115294555B (en) | 2022-12-20 |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| CN108537742B (en) | Remote sensing image panchromatic sharpening method based on generation countermeasure network | |
| Aich et al. | Deepwheat: Estimating phenotypic traits from crop images with deep learning | |
| Hayes et al. | Comparison of change-detection techniques for monitoring tropical forest clearing and vegetation regrowth in a time series | |
| Funke et al. | Efficient automatic 3D-reconstruction of branching neurons from EM data | |
| Millan et al. | On‐the‐Go Grapevine Yield Estimation Using Image Analysis and Boolean Model | |
| CN112561876B (en) | Image-based water quality detection method and system for ponds and reservoirs | |
| CN114283158A (en) | Retinal blood vessel image segmentation method and device and computer equipment | |
| CN114241340A (en) | Image target detection method and system based on double-path depth residual error network | |
| de Silva et al. | Towards agricultural autonomy: crop row detection under varying field conditions using deep learning | |
| CN115294555B (en) | Plant disease intelligent diagnosis method and system based on neural network | |
| CN110288574A (en) | A kind of adjuvant Ultrasonographic Diagnosis hepatoncus system and method | |
| CN113837154B (en) | Open set filtering system and method based on multitask assistance | |
| Wang et al. | An efficient attention module for instance segmentation network in pest monitoring | |
| CN114119532A (en) | Building change detection method based on remote sensing image and twin neural network | |
| CN112767331A (en) | Image anomaly detection method based on zero sample learning | |
| CN117437691A (en) | Real-time multi-person abnormal behavior identification method and system based on lightweight network | |
| CN112633301A (en) | Traditional Chinese medicine tongue image greasy feature classification method based on depth metric learning | |
| CN114596244A (en) | Infrared image identification method and system based on visual processing and multi-feature fusion | |
| CN117115675A (en) | Cross-time-phase light-weight spatial spectrum feature fusion hyperspectral change detection method, system, equipment and medium | |
| Li et al. | Image processing for crop/weed discrimination in fields with high weed pressure | |
| CN116091492A (en) | Image change pixel level detection method and system | |
| CN113221731B (en) | Multi-scale remote sensing image target detection method and system | |
| Song et al. | Detection of Northern Corn Leaf Blight Disease in Real Environment Using Optimized YOLOv3 | |
| KR102576427B1 (en) | Real-time Rainfall Prediction Device using Cloud Images, and Rainfall Prediction Method using the same, and a computer-readable storage medium | |
| CN114529794A (en) | Infrared and visible light image fusion method, system and medium |
Legal Events
| Date | Code | Title | Description |
|---|---|---|---|
| PB01 | Publication | ||
| PB01 | Publication | ||
| SE01 | Entry into force of request for substantive examination | ||
| SE01 | Entry into force of request for substantive examination | ||
| GR01 | Patent grant | ||
| GR01 | Patent grant |