ENHANCEMENT ANALYSIS OF IMMUNE FLUORESCENT CELL IMAGES

There are different patterns of immune fluorescence cells, which serve in determining different autoimmune disease. Hence, clearly identifying the features of the figures in the image will assist in automating the classification of these patterns. This project aims to enhance the quality of the Hep2-cell images obtained from Indirect Immune Fluorescence (IIF) Test. The enhancement of the quality in this project will be focused on enhancing the contrast, reducing the noise, and sharpening the edges of images. This enhancement will have a real serious impact on the stages coming after, which are patterns recognition and automatic classification. Creating an automatic battern classification system will improve the diagnostic process of the autoimmune disease instead of handling it manually. Consequently, many disadvantages of the manual interpretation can be overcome, such as level of expertise, time consuming and prone to mistakes. This research analyzed the performance of three enhancement approaches namely wavelet transform filter, diffusion filter, and wavelet transform filter combined with diffusion filter. The combination of wavelet transform filter with diffusion filter produced better result. However, the diffusion filter produced best result among all the three enhancement approach of the indirect immune fluorescence images. The recommendation for the future work is to explore an automatic determination of noise variance in the image when wavelet transform filter is being applied.