Plant Leaf Disease Detection and Classification using Optimized CNN Model
Prabavathi S1, Kanmani P2
1Prabavathi S*, PG Scholar, Department of Computer Science and Engineering, Christ (Deemed to be University), Bangalore, India.
2Kanmani P, Assistant Professor, Department of Computer Science and Engineering, Christ (Deemed to be University), Bangalore, India.
Manuscript received on March 20, 2021. | Revised Manuscript received on March 26, 2021. | Manuscript published on March 30, 2021. | PP: 233-238 | Volume-9 Issue-6, March 2021. | Retrieval Number: 100.1/ijrte.F5572039621 | DOI: 10.35940/ijrte.F5572.039621
Open Access | Ethics and Policies | Cite | Mendeley
© The Authors. Blue Eyes Intelligence Engineering and Sciences Publication (BEIESP). This is an open access article under the CC BY-NC-ND license (http://creativecommons.org/licenses/by-nc-nd/4.0/)
Abstract: Our economy depends on productivity in agriculture. The quantity and quality of the yield is greatly affected by various hazardous diseases. Early-stage detection of plant disease will be very helpful to prevent severe damage. Automatic systems to detect the changes in the plants by monitoring the abnormal symptoms in its growth will be more beneficial for the farmers. This paper presents a system for automatic prediction and classification of plant leaf diseases. The survey on various diseases classification techniques that can be used for plant leaf disease detection are also discussed. The proposed system will define the cropped image of a plant through image processing and feature extraction algorithms. Enhanced CNN model is designed and applied for about 20,600 images are collected as a dataset. Optimization is done to enhance the accuracy in the system prediction and to show the improvement in the true positive samples classification. The proposed system shows the improvement in the accuracy of prediction as 93.18% for three different species with twelve different diseases.
Keywords: Agriculture, Classification, CNN, Image processing, Optimization.