Description:FIELD OF THE INVENTION
This invention relates to Vision Enhanced and Noise Destroy Generative Adversarial Network for Improving Deep Learning Based Tomato Leaf Diseases Prediction
BACKGROUND OF THE INVENTION
Early and accurate prediction of tomato plant diseases remains a major challenge in modern agriculture, particularly in areas with limited resource settings. Deep Learning models provides efficient result in diseases prediction task, but they are often struggled by the limited sample size and data imbalance leading to poor diseases recognition and classification. Data augmentation is applied to artificially increase the amount of data by generating new data points from existing data. CycleGAN is widely used augmentation technique which can perform image translation on an unpaired image where there is no relation between input and output images. But, it suffers from relatively slow rate of convergence and delaying model optimization. Also, translation low- resolution images often lead to poor-quality results and significantly lowering the model performances.
Conventional methods for tomato plant diseases detection primarily rely on manual inception which are usually time-consuming, labor-intensive and prone to human-error especially in large-scale or limited resources farming environments. To solve this, AI based models are introduced, but the limitations like small data size, poor-quality images and class imbalance issues leads to sub-optimal performances in diseases classification. A common data augmentation model CycleGAN, provides image translation using unpaired datasets but suffers from slow convergence and low-resolution outputs when input images are of poor quality. The proposed system overcomes these limitations through a novel dual-generator architecture, VEND-GAN which simultaneously reduces image noise and enhances image resolution using MAM and RRDB modules. These modifications significantly reduce the constraints in traditional GAN models. Combined with PCA for efficient feature extraction and FRCNN for precise disease detection and localization, Moreover, its modular design ensures cost-effective scalability, making it well-suited for real-world agricultural deployment where resources and infrastructure maybe limited. This approach transforms traditional practices into an intelligent, automated and high-accuracy tomato plant disease detection framework.
SUMMARY OF THE INVENTION
This summary is provided to introduce a selection of concepts, in a simplified format, that are further described in the detailed description of the invention.
This summary is neither intended to identify key or essential inventive concepts of the invention and nor is it intended for determining the scope of the invention.
To further clarify advantages and features of the present invention, a more particular description of the invention will be rendered by reference to specific embodiments thereof, which is illustrated in the appended drawings. It is appreciated that these drawings depict only typical embodiments of the invention and are therefore not to be considered limiting of its scope. The invention will be described and explained with additional specificity and detail with the accompanying drawings.
The proposed model addresses the challenges in tomato diseases detection like limited data availability, poor image quality and slow training convergence. It introduces a novel image augmentation method using Vision Enhanced and Noise Destroy GAN (VEND-GAN) model which utilizes two generators. The first generator applies Mutual Attention Mechanism (MAM) for noise reduction, while second generator utilizes Residual in Residual Dense Block (RRDB) for image resolution enhancements. These two generators integrate to produce high-quality, noise-free images from low resolution dataset. A single discriminator is applied to distinguish the real and generated images which are guided by adversial and feature loss. The generate images are processed by Principal Component Analysis (PCA) for extracting relevant features, Faster Region Convolutional Neural Network (FRCNN) for precise diseases detection and localization. This model improves the classification accuracy making it suitable for various agricultural environments.
BRIEF DESCRIPTION OF THE DRAWINGS
The illustrated embodiments of the subject matter will be understood by reference to the drawings, wherein like parts are designated by like numerals throughout. The following description is intended only by way of example, and simply illustrates certain selected embodiments of devices, systems, and methods that are consistent with the subject matter as claimed herein, wherein:
FIGURE 1: SYSTEM ARCHITECTURE
The figures depict embodiments of the present subject matter for the purposes of illustration only. A person skilled in the art will easily recognize from the following description that alternative embodiments of the structures and methods illustrated herein may be employed without departing from the principles of the disclosure described herein.
DETAILED DESCRIPTION OF THE INVENTION
The detailed description of various exemplary embodiments of the disclosure is described herein with reference to the accompanying drawings. It should be noted that the embodiments are described herein in such details as to clearly communicate the disclosure. However, the amount of details provided herein is not intended to limit the anticipated variations of embodiments; on the contrary, the intention is to cover all modifications, equivalents, and alternatives falling within the scope of the present disclosure as defined by the appended claims.
It is also to be understood that various arrangements may be devised that, although not explicitly described or shown herein, embody the principles of the present disclosure. Moreover, all statements herein reciting principles, aspects, and embodiments of the present disclosure, as well as specific examples, are intended to encompass equivalents thereof.
The terminology used herein is for the purpose of describing particular embodiments only and is not intended to be limiting of example embodiments. As used herein, the singular forms “a",” “an” and “the” are intended to include the plural forms as well, unless the context clearly indicates otherwise. It will be further understood that the terms “comprises,” “comprising,” “includes” and/or “including,” when used herein, specify the presence of stated features, integers, steps, operations, elements and/or components, but do not preclude the presence or addition of one or more other features, integers, steps, operations, elements, components and/or groups thereof.
It should also be noted that in some alternative implementations, the functions/acts noted may occur out of the order noted in the figures. For example, two figures shown in succession may, in fact, be executed concurrently or may sometimes be executed in the reverse order, depending upon the functionality/acts involved.
In addition, the descriptions of "first", "second", “third”, and the like in the present invention are used for the purpose of description only, and are not to be construed as indicating or implying their relative importance or implicitly indicating the number of technical features indicated. Thus, features defining "first" and "second" may include at least one of the features, either explicitly or implicitly.
Unless otherwise defined, all terms (including technical and scientific terms) used herein have the same meaning as commonly understood by one of ordinary skill in the art to which example embodiments belong. It will be further understood that terms, e.g., those defined in commonly used dictionaries, should be interpreted as having a meaning that is consistent with their meaning in the context of the relevant art and will not be interpreted in an idealized or overly formal sense unless expressly so defined herein.
The proposed model addresses the challenges in tomato diseases detection like limited data availability, poor image quality and slow training convergence. It introduces a novel image augmentation method using Vision Enhanced and Noise Destroy GAN (VEND-GAN) model which utilizes two generators. The first generator applies Mutual Attention Mechanism (MAM) for noise reduction, while second generator utilizes Residual in Residual Dense Block (RRDB) for image resolution enhancements. These two generators integrate to produce high-quality, noise-free images from low resolution dataset. A single discriminator is applied to distinguish the real and generated images which are guided by adversial and feature loss. The generate images are processed by Principal Component Analysis (PCA) for extracting relevant features, Faster Region Convolutional Neural Network (FRCNN) for precise diseases detection and localization. This model improves the classification accuracy making it suitable for various agricultural environments.
NOVELTY:
The proposed work addresses the limitation of traditional GAN models by developing dual- generators to produce low-noise and enhance images for tomato disease prediction. The proposed GAN model (VEND-GAN) enables the high-quality image augmentation for low- resolution and unpaired dataset, overcoming the drawback of traditional GAN models. Generator 1 applies MAM that effectively handles complex noise thereby improving the precise restoration of the actual image content. Generator 2 utilizes RRDB to enhance the image resolution by capturing the fine details and textures of tomato plants. Also, it enhances the gradient flow of image features and simplifies training by focusing on input-output differences enriching representation across multiple scales. The generated Noise Reduction and High Resolution (NRHR) images from combined generator1 and genrator2 are input into a single discriminator that distinguishes real from fake images. Also, the adversarial loss is utilized to update the discriminator's weights during the training task and feature loss ensures the generated output aligned with the target image to preserve key characteristics and facilitate meaningful feature learning. Finally, the generated features are input into PCA for feature extraction, followed by FRCNNfordiseaselocalizationanddetection.Thispipelineboostsclassificationaccuracy,shortens training time and removes the need for large labeled datasets offering a cost-effective and scalable solution for real-world agricultural use in tomato plant diseases prediction.
, Claims:VISION ENHANCED AND NOISE DESTROY GENERATIVE ADVERSARIAL NETWORK FOR IMPROVING DEEP LEARNING BASED TOMATO LEAF DISEASES PREDICTION
The present invention relates to a Vision Enhanced and Noise Destroy Generative Adversarial Network (VEND-GAN) designed to improve deep learning-based tomato leaf disease prediction. The invention addresses key challenges such as limited data availability, poor image quality, and slow training convergence. The proposed VEND-GAN model comprises two generators: the first utilizes a Mutual Attention Mechanism (MAM) for effective noise reduction, while the second employs a Residual in Residual Dense Block (RRDB) structure to enhance image resolution. The dual generators work in tandem to produce high-quality, noise-free images from low-resolution datasets. A single discriminator is used to distinguish between real and generated images, guided by adversarial and feature loss. The generated images undergo Principal Component Analysis (PCA) for relevant feature extraction, followed by processing through a Faster Region-based Convolutional Neural Network (FRCNN) for precise disease detection and localization. The invention enhances classification accuracy, reduces training time, and minimizes the need for large labeled datasets, offering a cost-effective and scalable solution for real-world agricultural applications in tomato plant disease prediction.