DEVELOP THE HERBS PLANT CLASSIFICATIONMODEL FOR PATIENT CARE USING DEEP LEARNING

SURAFEL, AMSALU TADESSE

dc.contributor.author	SURAFEL, AMSALU TADESSE
dc.date.accessioned	2024-04-19T08:30:17Z
dc.date.available	2024-04-19T08:30:17Z
dc.date.issued	2023-06
dc.identifier.uri	http://ir.bdu.edu.et/handle/123456789/15770
dc.description.abstract	This paper focuses on noise removal in the classification and description of herbal plants using the Xception model. The study aims to improve accuracy by addressing challenges related to shadow removal, dust removal, and color correction through advanced algorithms. The research incorporates the Shadow Detection and Removal (SDR) algorithm for shadow removal, the Top-hat transform algorithm for dust removal, and the Gray World Algorithm with Contrast Limited Adaptive Histogram Equalization (CLAHE) for color correction and histogram enhancement. To build a diverse and extensive dataset, the data is collected from Kaggle, a popular platform for machine learning datasets. Extensive model training and optimization are performed using the powerful Xception model for feature extraction and classification. By integrating noise removal algorithms, including SDR, Top-hat transform, Gray World Algorithm, and CLAHE, the model achieves improved accuracy by effectively handling shadow, dust, and color inconsistencies. These algorithms enhance input data quality and enable more precise feature extraction. Our evaluation demonstrates exceptional accuracy, achieving a classification accuracy of 99%. Successful noise removal, encompassing shadows, dust, and color variations, minimizes erroneous classifications, enhancing the overall reliability of the model. Additionally, during the testing phase, the model demonstrates strong performance with a test accuracy of 97.5%, providing consistent and reliable results. Keywords: SDR, Top-hat transform, Gray World Algorithm, and CLAHE, Xception model	en_US
dc.language.iso	en_US	en_US
dc.subject	Software Engineering	en_US
dc.title	DEVELOP THE HERBS PLANT CLASSIFICATIONMODEL FOR PATIENT CARE USING DEEP LEARNING	en_US
dc.type	Thesis	en_US