Applying Smart System for Object Color Classification Using Color Average Technique

Witthaya Boonsuk; Yodrak Saisin

Witthaya Boonsuk Department of Information Technology, Nakhon Phanom University, Nakhon Phanom, Thailand
Yodrak Saisin Department of Information Technology, Nakhon Phanom University, Nakhon Phanom, Thailand

Keywords: Comparison, Image Proceeding, Algorithms, Digital Image, Color Intensity

Abstract

The objective of this research was to design an object color classification algorithm. The system design principle utilizes the color average technique consisting of 3 color groups: red, blue, and green. These colors were employed in object color classification. The results found that the efficiency assessment of the system was based on a newly developed algorithm from 3 sample groups with 30 images per group, which are 90 images in total. The images were 640x480 pixels in quality with average of 95.5% precision. The precision of group 1: Red had 96.6% precision; Group 2: Blue had 96.6% precision and Group 3: Green color had 93.3 % precision. The total average of the system was considered as high precision, which indicated the quality of the color group. Moreover, the results from this experiment can be applied in color sorting of objects.