ABSTRACT In this paper, a novel approach is made to discard the degradations in full shoe prints and partial shoe prints, in processing those images for recognition. A pass band DCT coefficient has been used to extract feature vectors. A more robust approach has been dealt with to find the matching between the partial shoe prints and the images in the data base. This method makes the shoe print recognition process more robust against degradations like noises, orientations and blurred images which are common in shoe print images and also helps in saving the processing time and memory consumption. General Terms Forensic analysis, Pattern Recognition and Artificial Intelligence Keywords Shoe print, Fisher Linear Discriminant (FLD), recognition, forensic science, Principal Component Analysis (PCA). 1. INTRODUCTION In forensic analysis, shoe marks provide valuable information as evidence against accused persons. In many cases, shoe prints can be positively identified as having been made by a specific shoe to the exclusion of all other shoes. The identification is based on the physical match of individual designs. The evidence provided by a positively identified shoe mark is as strong as the evidence from fingerprints, tool marks, and typewritten impressions [1]. But this matching alone may not be the only criteria to identify the suspected person in crime scene. But, still this information is a valuable one. Due to the infinite designs available in the market, with distinctive outsole patterns, this implies that there exists specific general population of shoe prints [1]. If the model of a shoe can be determined from its mark, then this can abruptly narrow the search for a particular suspect. So, at least the forensic department can narrow down their search of suspected persons. The data base formation is the first step in shoe print processing. So matching algorithm starts with image acquisition process. An image of a shoe print is first obtained using a technique such as photography, electrostatic lifting or by making a cast when the impression is in soil, snow or sand. Further, in the forensic laboratories, the image of the shoe mark is compared with the shoeprints and shoe impressions of known shoe samples. A process of detection and recovery of footwear impression evidence and of comparison of the impressions with suspect shoes has been explained in [1]. The photograph of the impression or of the lifted impression or cast can be subsequently scanned and a digital image is produced. Forensic analysis requires comparison of this image against specific databases. These databases include, database of prints made by shoes of already existing criminals and previously available on the market and database of shoe prints found at other crime scenes. Comparing crime scene shoe mark images to databases is currently a difficult task and it is usually conducted manually by searching paper catalogues or computer databases. Our task is to simplify the matching process against degradations and help the forensic scientists for faster search.