Comparison on Different Filters for Performance Improvement on Fingerprint Image Enhancement

Today, in the era of Internet, biometric authentication plays many important roles. And for biometric authentication image, enhancement plays a vital role because the recognition of an image is extremely depend on it. The image quality can be improved using image enhancement phase after removing noise, joining damaged ridges and constructing smooth images. Various filtering techniques for image enhancement like min filter, max filter, median filter, mean filter, Gabor filter, O’Gorman filter as well as segmentation techniques like Sobel edge detection are used. In the proposed research work, above all mentioned filters are applied and compared the result for enhancing the fingerprint images. The comparison of all filters is done based on MSE, PSNR and computational time to measure the performance of the proposed research work. The implementation of the proposed research work was done using Java and used custom database of 100 images of 25 users.