Ink Mismatch Detection using Unsupervised learning
Usman Zafar, Raqiyya BiBi · 2023
Hyperspectral imaging has emerged as a powerful technique for capturing detailed spectral information in document analysis application.For ink mismatch analysis in a handwritten document hyperspectral document images offer additional information to identify subtle variations in ink color.Such information is not available in normal images.In this work kmeans clustering is used to detect different ink in hyperspectral document image.At first hyperspectral data is preprocessed to reduce noise and background.Subsequently kmeans clustering is applied to group the data into different clusters