Green pepper recognition based on color and texture characteristics
In order to improve the recognition efficiency and picking accuracy of the green pepper picking robot on the green pepper fruit in the natural environment,a green pepper recognition method based on the color and texture characteristics is proposed,and the conventional RGB color camera can achieve a better recognition effect in the natural environment by using the conventional RGB color camera.Firstly,the green pepper image is converted from RGB to HSV color space,and after comparative analysis of the color difference of the S-V component,the fruit can be highlighted,the complex background can be removed,and then the LBP features and HOG features of the green pepper are extracted,a single feature model and a multi-feature fusion model are established,and different classifier SVM and AdaBoost are used for feature training to find out the classification algorithm that is most suitable for green pepper recognition.Experimental results show that the recognition accuracy of LBP+HOG+AdaBoost algorithm reaches 99.3％,which is better than other models.This study can provide a research basis for the intelligent identification of green pepper picking machines.
green pepper recognitionSVMAdaBoostcolor characteristicstexture characteristics
万方数据
维普
