Algoritma pengolahan citra tanaman cabai merah menggunakan pengindera tiga dimensi

Abstract

Machine vision is one of the image processing techniques used to process the image of an object according to its color, shape, size by using a video camera as an image capturer. The image processing used in this study, could be categorized as one cf the machine vision techniques, eventhough using three dimensional vision sensor as an image capturer. The three dimensional vision sensor was constructed using two laser beams namely the red laser (685 nm) and the infrared laser (830 nm) as the light transmitter and two position sensitive devices (PSDs) as the receiver unit. The output signals of each PSD were detected from two output terminals called anode A and anode 8. Three kind of signals namely red A+B, infrared A and infrared B were collected from both PSDs. This vision sensor was constructed with the aims to recognize the red ripe chili fruits from their canopy and to calculate the cutting point of the individual fruit's stem three dimensionally. The ripeness of the red chili fruits were analyzed using the ratio between red A+B and infrared A+B while the distance of each picture element (pixel) to the center coordinate of the vision was calculated using the ratio between infrared A and infrared A+B. The center point of the image was calculated using the combination of the layering or chain technique and the moment area technique. The results of the experiment showed that about 82 % of the red ripe fruit were successfully detected from 20 samples of the chili plant that were containing 39 red ripe fruits. Three-dimensional location of the center image and the cutting point of the red ripe fruit stem was detected with accuracy of about 10 mm.