This research illustrates methods based on computer vision for the recognition of characters in Gujarati Sign Language. This study is for Gujarati Sign Language Character Recognition using Shape, Texture, Zone and Hand Finger feature sets. In this research, 294 images of Gujarati Sign Language characters are used, and the images are classified using Neural Network and Multi-class Support Vector Machine classifiers. Proposed system models are trained using 205 images of Gujarati Sign Language characters ક (43), ખ (42), ચ (37), છ (43), and જ (40), while system models are tested using 89 images of Gujarati Sign Language characters, The system models are tested using 89 images of Gujarati Sign Language characters ક (20), ખ (21), ચ (15), છ (15), and જ (18). Results found that the Neural Network obtained an accuracy of 95.50%, while the Multi-Class Support Vector Machine obtained an accuracy of 85.39%. Classification Report, Confusion Matrix, Calibration Curve, and ROC analysis diagram criteria are used to evaluate classifier performance. Neural Networks have been found to perform better than Multi Class Support Vector Machine classifiers.