An Application Of Fuzzy Mathematics In Color Identification Of Robots

For the influence of environment and hardware, the recognition rate of color is usually not good for robots. This paper presents a resolution by means of an effective and simple algorithm based on fuzzy theory. The method extracts the color components of image in RGB color space into the color space associated I 1,I₂ and I₃with a dynamic K-L transformation. It calculates the membership degree, induces and improves the degrees respectively, then recognizes the color by three color components of image based on comprehensive judge principle of fuzzy sets theory. This method can be used to simulate real-time color identification in lower hardware environment. Because it obtains the related color information of image directly and deals with these data based on fuzzy theory, the results of identification must have not been affected by the subjective discrimination and the speed has been increased obviously. The experimental data shows clearly it is a simple and effective method.