Sign language is the language of visual gestures that are mainly used as a communication tool by deaf community. Sign languages use visual pattern that are used to communicate rather than acoustic patterns that are used in verbal communication. Sign language can be a benchmark for gesture recognition system as it is the most structured and developed form of gestures. Automated Sign Language Recognition (SLR) has very effective uses in many real world domains. There are many applications of SLR in the field of robot control, interactive learning, appliances control, virtual reality, simulations, games, industrial machine control, and many more apart from its significance for hearing impaired community. Sign language is not an international language as sign languages are not uniform throughout the world. Like verbal languages, sign languages also differ from region to region and country to country. Pakistani Sign Language (PSL) is a visual-gestural language that came out as a blend of urdu, national language of Pakistan, and other regional languages. The thesis presents a novel, robust, reliable, systematic and consistent system for static PSL recognition. The thesis is based on the empirical evaluation of different potential sign descriptors. The pragmatic approach has lead to a mathematical sign model that has given convincing performance for PSL recognition in terms of accuracy. The polynomial parameterization is proposed as the sign model for PSL recognition. The inherent uncertainty of the domain of sign language demands a classification tool that respects this uncertainty. Because of this very reason, the fuzzy inference got the prominent lead when experimentally compared with other competing classifiers. The main contributions of the thesis are: the development of PSL dataset, robust and efficient sign descriptor and a fuzzy rule based inference model as classifier. There is no standard dataset available for PSL, so dataset for a subset of static signs of PSL is developed for the thesis. An empirical mathematical sign model is presented that has shown its supremacy when analyzed in comparison with other potential sign descriptors. This mathematical model defines every sign of xii the PSL dataset as a polynomial parametric model. For the classification of an uncertain domain like SLR, the conventional classifiers could not come up with sound results. So a fuzzy rule base is proposed for PSL recognition based on polynomial parameters of every individual sign. The meticulous statistical analysis of the proposed PSL Fuzzy Model (PSL-FM) has shown very convincing results.