Automatic recognition of mature fruits in a complex agricultural environment is still a\nchallenge for an autonomous harvesting robot due to various disturbances existing in the background\nof the image. The bottleneck to robust fruit recognition is reducing influence from two main\ndisturbances: illumination and overlapping. In order to recognize the tomato in the tree canopy\nusing a low-cost camera, a robust tomato recognition algorithm based on multiple feature images and\nimage fusion was studied in this paper. Firstly, two novel feature images, the a*-component image\nand the I-component image, were extracted from the L*a*b* color space and luminance, in-phase,\nquadrature-phase (YIQ) color space, respectively. Secondly, wavelet transformation was adopted to\nfuse the two feature images at the pixel level, which combined the feature information of the two\nsource images. Thirdly, in order to segment the target tomato from the background, an adaptive\nthreshold algorithm was used to get the optimal threshold. The final segmentation result was\nprocessed by morphology operation to reduce a small amount of noise. In the detection tests, 93%\ntarget tomatoes were recognized out of 200 overall samples. It indicates that the proposed tomato\nrecognition method is available for robotic tomato harvesting in the uncontrolled environment with\nlow cost.
Loading....