关于动态三维图像真实性识别仿真研究  

Simulation Research on Dynamic 3D Image Reality Recognition

作　　者：钱乐旦 周潇 

机构地区：[1]温州大学数学与信息科学学院,浙江温州325000 [2]温州大学现代教育技术中心,浙江温州325000

出　　处：《计算机仿真》2017年第12期361-365,共5页Computer Simulation

基　　金：浙江省教育厅科研项目(Y201327503)

摘　　要：对动态三维图像的真实性识别,能够有效提升高噪声图像的视觉质量。对三维图像真实性识别,首先需要增强图像平滑度,将三维图像像素点进行分类划分,通过细节区域窗口面积比例来融合各个单尺度增强结果,完成对三维图像真实性的识别。传统方法将深度信息与多尺度小波分解相结合,将分解获得的小波高频子设定人为操控因子,但忽略了融合图像各个单尺度增强结果,导致增强效果不理想。提出基于边缘检测加权引导滤波的三维图像真实性识别方法。计算三维图像的边缘算子幅值,分别获得了图像纹理区域和平滑区域的自适应归一化因子,利用模糊同组技术将三维图像像素点进行分类划分,通过细节区域和平滑区域所占局部窗口面积比例,融合三维图像各个单尺度增强结果,减少图像质量颜色失真。实验结果表明,所提方法能够有效突出三维图像质量细节,减少图像质量颜色失真,识别效果优越。In this paper, we present an authenticity identification method for three - dimension image of weighted guided filtering based on edge detection. Firstly, the amplitude of edge operator for three - dimension image is calculated to obtain adaptive normalization factor of image texture region and smooth region, and then three - dimension image pixels are classified by using fuzzy group technology. At last, enhancement result of each scale of three - dimension image is fused according to proportion of detail area and smooth area occupied by local window area, thus color distortion of image quality is reduced. The simulation results show that this method can effectively highlight quality detail of three - dimension image and reduce color distortion of image quality, and has excellent identification effect.

关 键 词：三维图像 真实性识别 平滑度增强 

分 类 号：TP391[自动化与计算机技术—计算机应用技术]