• Linigerew Mengstie Shita
• Prof.Mogalla Shashi

Coffee is the majorly traded commodity used by one–third of the world’s population as a beverage and Ethiopia is the home land of coffee arabica, it covers 7-10 % of the world’s coffee production. This research work, focuses on the four major types of coffee leaf diseases that reduces coffee production in Ethiopia ,these are brown eye spot (BES),coffee berry disease (CBD),coffee leaf rust (CLR) and coffee wilt disease (CWD).In this paper ,we proposed Ethiopian coffee leaf diseases identification using deep learning features. The images of the coffee leaf diseases were captured from the regions of Ethiopia where more coffee is produced, i.e. Jimma and Zegie. We compared gaussian filtering, median filtering and the hybrid of the two filtering techniques to remove noises from coffee leaf images and we have got better result from the hybrid of the two filtering techniques. And also we applied KMeans clustering for segmentation and CNN for feature extraction. Finally, we made a comparison between CNN-Softmax classifier and CNN-SVM classifier .The experimental results showed that SVM performs better than softmax classifier in terms of performance and computational time. Our proposed model with SVM classifier achieved an overall classification accuracy of 96.5%.

Ethiopian coffee, SVM, CNN, coffee disease

"ETHIOPIAN COFFEE LEAF DISEASES IDENTIFICATION USING DEEP LEARNING FEATURES", International Journal of Emerging Technologies and Innovative Research (www.jetir.org), ISSN:2349-5162, Vol.8, Issue 10, page no.b498-b508, October-2021, Available :http://www.jetir.org/papers/JETIR2110156.pdf

"ETHIOPIAN COFFEE LEAF DISEASES IDENTIFICATION USING DEEP LEARNING FEATURES", International Journal of Emerging Technologies and Innovative Research (www.jetir.org | UGC and issn Approved), ISSN:2349-5162, Vol.8, Issue 10, page no. ppb498-b508, October-2021, Available at : http://www.jetir.org/papers/JETIR2110156.pdf