Fingerprint has been used as a biometric feature for security reasons for centuries. Automated Fingerprint Identification System (AFIS) is one such authentication method used in wide range of application domains such as ecommerce and automated banking. Fingerprint image contains flow-like pattern called ridges which are separated by furrows. Ridge ending and bifurcation are two type of minutiaes used as basic features in AFIS. There are two approaches for minutiae extraction, namely conventional and direct. In the conventional approach, fingerprint images have to go through several processes including noise removal, enhancement, directional image computation, segmentation and thinning. Whereas, in the direct approach, the minutiaes are directly extracted from a gray scale image without going through all the above processes. Extracting minutiaes have been found to be an error prone process, depending on the quality of the fingerprint image. In the conventional approach, for instance, a low quality image will generate many artificial minutiaes which lead to errors in fingerprint matching. Similarly, in the direct approach, a bad quality image that contains scars, sweat spots and uneven ridges and furrows can lead to artificial minutiaes. This thesis presents a fingerprint image reconstruction algorithm using Directional Fourier filtering. Prior to the image reconstruction, a directional image is first computed using Mehtre technique and followed by a 3-tier enhancement processes viz. Histogram Equalization, High-pass filter and Median filter. By using the directional image as ridges orientations map, its original fingerprint image is filtered using the Directional Fourier filtering to produce a new fingerprint image. The reconstruction algorithm was tested with 500 fingerprint images. The results of the experiment is very promising.