Abstarct: The feature reduction increases the speed of processing and reduces the storage capabilities. Therefore, there has been a lot of attention in feature selection and feature extraction methods. In recent years Artificial Immune System algorithm (AIS) baxsed on clonal selection with its optimization and evolutionary properties highly regarded. By choosing appropriate method for calculating affinity and perform hypermutation and generating clones proportional to result of antibody and antigen affinity, the AIS algorithm has been achieved to acceptable results efficiently. In this thesis a feature vector consist of color properties in HSV channel and texture features calculated by Haralick method, extraced from each pixel of a picture that taken by mobile robot. For determining the road and non-road pixel the Support Vector Machine (SVM) method has been applied. AIS algorithm has the potential to generate both the optimal feature subset and SVM parameters at the same time. Our research objective is to optimize the parameters and feature subset simultaneously, without degrading the SVM classification accuracy. Clonal selection in AIS algorithm have high time complexity thus to reduce it, the adaptive clonal selection algorithm is used. Several public datasets of UCI repository are employed to calculate the classification accuracy rate. The experimental results conducted in this study with AIS algorithm baxsed on clonal selection algorithm and SVM as a function for affinity calculation, show that the average feature reduction rate is 68.62% with 94.26 average accuracy rate and for dataset extracted from mobile robot image, average feature reduction rate is 64.90% with 83.97 average accuracy rate.