1. The document describes a system for online signature verification using Discrete Cosine Transform (DCT). Signatures are acquired using a tablet and features are extracted from the x and y coordinates and theta values. DCT is applied to compress the signature features. 2. The system is tested on 125 signatures from 25 signers. Genetic algorithms are used for classification. The system provides output as "genuine signature" or "forged signature" and calculates false acceptance and rejection rates to evaluate performance. 3. Experimental results found the proposed DCT-based approach to be promising for online signature verification by extracting dynamic features while keeping basic signature information.