Рақамли технологияларнинг назарий ва амалий масалалари Том 8 № 4 (2025) · с. 71-78
Classification of Pneumonia based on a Concatenated CNN
Бурибоев, А.Ш.
Аннотация
In this study, a concatenated CNN model for pneumonia detection is proposed, combined with an image enhancement method based on fuzzy logic. The image enhancement process utilizes a novel refined fuzzification algorithm, which significantly improves image quality and feature extraction efficiency for the CCNN model. Four datasets were used for training: original images and images processed using fuzzy entropy, standard deviation, and histogram equalization. The results showed that using enhanced datasets notably increased the CCNN’s performance, with the fuzzy entropy–enhanced dataset yielding the best results. The proposed CCNN model achieved outstanding classification metrics, including 98.9% accuracy, 99.3% precision, 99.8% F1-score, and 99.6% recall. Experimental comparisons demonstrated that fuzzy logic–based image enhancement significantly outperforms traditional methods, providing higher diagnostic accuracy. This research highlights the effectiveness of integrating deep learning models with advanced image enhancement techniques for medical image analysis.
классификациянечеткая логикафаззификацияданныеконкатенированная CNNclassificationfuzzy logicfuzzificationdataconcatenated CNN
Источник метаданных: OAI-PMH архив журнала · Sindex не хранит полный текст, а даёт ссылку на источник.