Segmentation method of COI for monitoring and prediction of the crop growth

We are interested in the prediction of agricultural production and are trying to utilize an image processing technology into monitoring and prediction of the crop growth. At first, we must extract and collect the crop information such as location, size, leap area index, canopy, and etc of the crop for this purpose. In this paper, we suggested the effective segmentation method of COI(Crop Of Interest) at horticulture greenhouse. The proposed method in compose of two steps. The first, a color image of the crop is segmented the green and non-green region. The second, a depth image of the crop is removed near crops as rear crops and both sides crops. In the experiments, our method shows the correct segmentation of COI.