期刊名称：Journal of Theoretical and Applied Information Technology
印刷版ISSN：1992-8645
电子版ISSN：1817-3195
出版年度：2017
卷号：95
期号：5
出版社：Journal of Theoretical and Applied
摘要：Detection of Alternaria blight disease in bitter gourd leaf can be done early by applying Image Processing Techniques. The growth of the plant is categorized into five stages. Images are acquired separately for healthy, diseased and test leaves. Histogram analysis is done at each stages for the identification of healthy and diseased leaves. The images are pre-processed and the colour characteristics are analysed using RGB, HSV and HSI model. The values of healthy and diseased leaves at each stages are stored. These values are used for checking with value of test leaves for the early detection of disease in the leaves of bitter gourd plant. This extensive experiment helps the early detection of the disease. Neural network pattern recognition tool is used to find the accuracy of results obtained for healthy and disease leaves at each stages.
