Tea Leaf Diseases Recognition using Neural Network Ensemble

Abstract

This paper proposes a tea leaf diseases recognizer (TLDR), an initiative to recognize diseases of the tea leaf. In TLDR, at first the image of the tea leaf is cropped, resized and converted to its threshold value in the image processing. Then feature extraction method is applied. Neural Network Ensemble (NNE) was used for pattern recognition. The extracted features are passed to the ANN along with the disease type and the ANN is trained. When a new image is uploaded into the system the most suitable match is found and the disease is returned. After going through the testing process 91 % of accuracy was found. The proposed solution would support the tea industry of Bangladesh to grow in the global market and also increase its tea production by minimizing the effect of tea leaf diseases.