Feature Extraction And Recognition Based On Machine Vision Application In Lotus Picking Robot
Shuping Tang,Dean Zhao,Weikuan Jia,Yu Chen,Wei Ji,Chengzhi Ruan
DOI: https://doi.org/10.1007/978-3-319-48357-3_46
2016-01-01
Abstract:Recently the picking technology of high value crops has become a new research hot spot, and the image segmentation and recognition are still the key link of fruit picking robot. In order to realize the lotus image recognition, this paper proposes a new feature extraction method combined with shape and color, and uses the K-Means clustering algorithm to get lotus recognition model. Before the feature extraction, the existing pulse coupled neural network segmentation algorithm, combined with morphological operation, is used to achieve nice segmentation image, including lotus, lotus flower, lotus leaf and stems. Then in the feature extraction processing, the chromatic aberration method and the moment invariant algorithm are selected to extract the color and shape features of the segmented images, in which principal component analysis algorithm is selected to reduce the dimension of the color and shape features to achieve principal components of lotus, lotus flower, lotus leaf and stems. In the experiment, K-Means clustering algorithm is used to get lotus recognition model and four clustering centers according to above principal components of training samples about lotus, lotus flower, lotus leaf and stems; then the testing experiment is applied to validate the recognition model. Experimental results shows that the correct recognition rate is 90.57 % about 53 testing samples of lotus, and the average recognition time is 0.0473 s, which further indicates that the feature extraction algorithm is applicable to lotus feature extraction, and K-Means algorithm is simple, reliable and feasible, providing a theoretical basis for positioning and picking of lotus harvest robot.