Grayscale histogram feature extraction algorithm and its application in medical image analysis
Yanan Hao, Yong Chai Tan, Vin Cent Taı, Xiaodong Zhang, Enping Wei · 2023
This paper applied grayscale histogram feature extraction to analyse a medical image of tuberculosis lungs. Five different edge detection methods, namely Roberts, Sobel, Prewitt, Canny, and LoG algorithms were used to realise the feature extraction of the medical image and their performace were compared. This study finds that Roberts and Prewitt algorithms were able to extract most information from the medical image, followed by Sobel and Canny algorithms, while LoG algorithm was able to extract the least information from the medical image. The results show that the features extracted by the algorithms can be used for pathological analysis.