基于HSV颜色空间的水下物体识别技术研究

Abstract

In order to identify underwater objectives, which need to be separated from the region, on the basis of the target that can be further analyzed. In view of that only the intensity information of reflected light is used in underwater TV monitoring system, a novel, simple and effective technique based on polarization characteristics of reflection laser to recognize underwater objects is reported. This method combines Stokes imaging technology with HSV color space theory. It can show polarization information of objects with the same color but different materials in HSV space: according to the independence of hue, saturation and brightness in HSV color space, the saturation represents the degree of polarization and the hue represents the orientation of the polarized fraction of the light. Both the saturation and hue vary with different materials. The polarized characteristics of reflection light provide the basis for distinguishing targets with different materials. Several kinds of materials were measured by the novel system on the linear polarization of the incident light. The light is incident on the dielectric reflector and metal objects simultaneously, the variations of polarization are correlated to the index of refraction of targets with the same angle of incidence. It was demonstrated that the objects in different material could be distinguished effectively according to the different colors in HSV space. The technique can increase the contrast between the target and background, and improve the identification resolution of underwater imaging. We therefore consider the technique for recognition of underwater object based on HSV color space highly promising for a range of different applications in well testing, as well as ocean exploration and other fields.