AN EFFECTIVE SEGMENTATION AND LIMO CLASSIFICATION FOR PADDY DISEASE DETECTION USING DEEP LEARNING

Abstract

Rice crop disease detection and its diagnosis methods are vitally important for the agriculture field to be sustainable. For that many researchers finding solutions to minimize or avoid the rice plant disease to take the best yield for formers. Because this disease led to a more than 38% yearly drop in paddy production. Due to a lack of awareness and digital knowledge in fast identifying and best remedy for rice crop diseases. In that, automated and artificial intelligence (AI) based rice crop disease detection and prevention method is a key research solution needed for the current agriculture field. The internet of things (IoT), has plenty of opportunities and contributing a vital role in wireless networks, especially in the last 15 years. Using IoT in the agriculture industry is growing up rapidly as it receives complex contextual information about water irrigation, crop disease detection, fertilizer utilization, and soil rate. Various crop disease detection methods need more accuracy and dimensionality corrections. Disease detection is indispensable for agriculture to be maintainable. Meantime automated rice plant disease detection systems also face various problems to detect diseases in the current situation. Regular machine learning based image-wise disease detection methods are following preprocessing input values, necessary feature extraction, image segmentation, and disease classifications steps. The proposed research work provides solutions for the above-mentioned problems and requirements with a novel approach which is the combination of the laurent series with intelligent multidimensional object optimization (LIMO classification framework) based on generative adversarial network (GAN) and swarm intelligence optimal classification through cognitive attribute selection (SIOC-CAS) to recognize various types of crop diseases in an agricultural field. Through this proposed research work IoT nodes are senses the values of a field, and crop, and the gathered information is shared with the processing unit with base station communication. Multi-objective and cognitive learning-based routing (MOCLEAR) protocol supports choosing optimal paths for data transmission betterment. After MOCLEAR protocol communication, receives crop data such as image then data reduction and dimensional sets creation to improve the eminence of the input values for segmentation and classification. Then, image segmentation using GAN combined with cognitive residual convolution network (CRCNet) is modified to segment values from input images. After receiving segment input images perform under feature iv extraction and image classification with significant attributes. Generally, a corpus of attributes is extracted to better characterize a pattern recognition problem. Attribute selection plays a vital role in image classification to the complexity of handling many features and increases the possibility of prediction rate. From a broader perspective, this stage greatly influences the recognition process, in terms of both predictive performance and the design of a computationally simple classifier.