The classification flesh aromatic coconuts in daylight

The purpose of this paper was to develop an efficient and accurate classification for flesh aromatic coconuts in daylight by using image processing technique. In actual implementation, the brightness of daylight was not constant as a major problem that is affecting for the classification flesh aromatic coconuts. So we need to adjust the brightness of image to have the same brightness in all images before to be classify flesh aromatic coconut. The color of the coconut's rind around the bottom of aromatic coconuts are correlated with coconuts age. Convert the region of interested in RGB color to HSV color then find the appropriately threshold and then draw 4 circular ring area into the region of interested. After that find the percentage of the white pixels of each ring and find relationship between the ring order and the percentage of the white pixels of each rings by using polynomial regression equation. The results showed that the system could classify flesh aromatic coconuts at 92.8% with single layer, 71.43% with one and half layer and 72.2% with double layer.