Aim at the color difference between the regular materiel and the foreign and randomicity of foreign color in fast foreign real-time recognition

the specialities of OC-SVM using in foreign recognition are studied

and an approach to real-time recognition based on hypersphere one-class support vector machines is presented. In this method

the color distributing of the regular materiel is obtained though OC-SVM

that is to find out the foreign whose color are not in the range of the regular materiel. To solve the OC-SVM

centrifugal coefficient ω is presented

and Zoutendijk fastest decline method is adopted to seek working set

so as to predigest SMO method. The result of experiment shows

this meshod can reduce about 20% CPU time than LibSVM; contrast to one dimension normal distribution and three dimension normal distribution fitting algorithm

the general accuracy of recognition is 8~10% higher by using this method

especially for some foreign whose color are similar to that of regular materiel.