Abstract: In order to guide the robots for harvesting citrus fruit, a method based on color image processing was developed to identify citrus fruit in the tree canopy. Fifty-three color images with natural citrus-grove scenes were analyzed using the range of interest(ROI) tool. And the color properties of target objects (citrus fruit, leaves, or branches) were extracted. The values of R subtracted by B of citrus fruit are different from those of leaves and branches. A rule for segmenting citrus fruit from background is put forward that the pixel belongs to citrus fruit if the value of R subtracted by B is larger than the threshold value (T), which is decided by dynamic threshold segmentation method, otherwise the pixel belongs to background. The results show that the visible fruits under direct sunlight or backlighting condition can be identified well using this method.