Research of Segmentation Method on Image of Lingwu Long Jujubes Based on a New Extraction Model of Hue

This paper studies on the image segmentation method of Lingwu Long Jujubes based on a new extraction model of hue to improve the accuracy of extracting images of Lingwu Long Jujubes. According to the characteristics that color components of Lingwu Long Jujubes in RGB color space have different distribution in a shadow environment or others, a extraction model of hue aiming at images of Lingwu Long Jujubes based on stage treatment of R component is built to extract hue information. And the difference between the target object and the background object is increased by using color difference. Then, the image is segmented by optimal threshold obtained by combining the maximum entropy and the mathematical criteria to achieve the adaptive adjustment of the segmentation threshold. Finally, the segmented image will be obtained through dealing with mathematical morphology. By comparing the segmentation effect of 30 Lingwu Long Jujubes images with artificial methods and other methods, it proves that the color image segmentation method of Lingwu Long Jujubes based on a new extraction model of hue has good effect to extract the object region. The accuracy of the segmentation rate is up to 92.6883%. The time that the algorithm run is 1.3107 s.