The Quick Recognition Method For Chinese Rose Based On Mobile Phone

The quick recognition of flower and plants is a hotspot in today's Horticultural Plants research.We chose the Chinese rose which lives in the natural background as a research object.Using the convenience and advantages of mobile phone,a quick recognition method for Chinese rose is proposed in this paper.Using the camera of the mobile device to obtain the original image of the rose flower,and then using the human-computer interactive segmentation method to preprocess the image,After Get the pre-processed image,the main color and shape features using the new main color extraction algorithm and the shape feature description algorithm are extracted respectively.Then the texture features are extracted by applying the improved fusion GLCM and LBP algorithm,and 22 features are obtained.Using Feature Selection Algorithm get 18 strong classification characteristics.Finally,Random Forest algorithm is used to test 1000 images,the average recognition rate can reach 93.5%.The result shows that the method has a good recognition effect and fast recognition speed.