An optimization on pictogram identification for the road-sign recognition task using SVMs

摘要: Pattern recognition methods are used in the final stage of a traffic sign detection and system, where main objective is to categorize detected sign. Support vector machines have been reported as good method achieve this target due their ability provide accuracy well being sparse methods. Nevertheless, for complete data sets signs number operations needed test phase still large, whereas needs be improved. The objectives work propose pre-processing improvements support increase achieved while vectors, thus phase, reduced. Results show that with proposed increased 3-5% reduction vectors 50-70%.