Method for detecting citrus fruits on tree

Abstract

The invention relates to a method for detecting citrus fruits on a tree. The method comprises the steps that the local and global information entropies are combined to analyze the illumination distribution characteristic of an image, and a local block homomorphic filtering algorithm is used to process illumination unevenness of the image; an adaptive enhanced RG color difference method is used toimprove the contrast between a foreground target and the background; threshold segmentation and a mathematical morphology method are used to realize the binarization of an RG chromatic aberration diagram; for the problem of multi-fruit adhesion or mutual occlusion in an acquired binary image, marker-controlled watershed transformation is used to realize adhesion region segmentation; a convex hulltheory is used to solve the problem of region under-segmentation caused by occlusion to acquire an independent candidate citrus fruit region; the histogram description feature of the candidate citrusfruit region is extracted; and citruses are identified by a histogram cross-core support vector machine to complete fruit location. According to the technical scheme provided by the invention, the accuracy and reliability of the detection of the citrus fruits on the tree are improved.

Description

technical field [0001] The invention relates to the fields of agricultural intelligence, precision agriculture, machine vision and pattern recognition, in particular to a method for detecting citrus fruits on trees. Background technique [0002] With the increase of orchard management costs and the decrease of skilled labor, the sustainability of traditional citrus fruit labor-intensive picking operations has shown a downward trend (Gongal A, Amatya S, Karkee M, et al. Sensors and systems for fruit detection and localization. Computers and Electronics in Agriculture, 2015, 116:8-19). A citrus picking robot is a kind of automatic operation equipment that changes the traditional way. One of the main problems it faces is the rapid and reliable detection of citrus fruits on the tree. [0003] Sensing technologies commonly used in fruit detection systems include visual (visible light) imaging, infrared imaging, multi-spectral / hyperspectral technology, binocular / multi-eye imaging...

AI Technical Summary

Problems solved by technology

However, the computational complexity of the above multi-scale sliding window method, Canny operator, Hough transform and SIFT algorithm is high, which will affect the real-time performance of fruit detection

[0005] In summary, although the current fruit detection method based on the monocular vision imaging system has achieved certain results, in natural scenes with changing lighting conditions, the fast and reliable detection of citrus fruits for the problems of multi-fruit adhesion and scene participant occlusion The method still needs to be further improved

Images