Implementation of starfruit maturity classification algorithm on embedded system

This paper presents an implementation of starfruit maturity classification algorithm in an embedded real-time system. In order to do classification for the starfruit, the maturity of the starfruit is classified based on new color feature, which is the hue between 2 color component of red and green denoted as C. Field Programmable Gates Array (FPGA) is selected to be the main processor of the system. Not only because of the capability of parallel execution on the algorithm, the FPGA is chosen also because of its low power consumption. The acquisition device for the system is a CMOS camera, which immune to high noise and static low power. From the CMOS camera, the image is transferred into the FPGA through ITU-R 656 Decoder. With YCbCr 4:2:2 color system, firstly the system will segment the starfruit image in order to remove the background. Next, index maturity classification is performed. In this work, classification accuracy of 90% was achieved. The index result is shown at Liquid Crystal Display (LCD).