工业产品造型数字化视觉质量评估优化仿真

Industrial Product Modeling Digital Visual Quality Assessment Optimization Simulation

针对当前视觉质量评估方法存在视觉机械图像分辨率低,评估时间慢,误差较大等问题。提出一种基于加权结构相似度视觉质量评估优化方法。利用子波分解方法对机械产品图像进行分解,得到机械产品图像的低频子带和高频子带,采用视觉模型对每个机械产品图像子带进行感知滤波处理,采用加权结构相似度方法进行评估,实验结果表明,所提方法评估时间较短,视觉图像分辨率较高,误差较小,能很好满足工业产品造型数字化视觉质量评估需求。

There are some problems about visual mechanical image, such as low resolution ratio, slow assessment process and large error in current methods of visual quality assessment. In this paper, a method to optimize visual quality assessment based on weighted structural similarity was presented. The wavelet decomposition method was used to decompose mechanical product image and obtain the low frequency sub-band and high frequency sub-band of mechanical product image. Moreover, the visual model was applied to the perceptual filtering process of each mechanical product image sub-band. Finally, the weighted structural similarity method was used for the evaluation. Simulation results prove that the proposed method has shorter assessment time, higher image resolution ratio and less error. Meanwhile, this method can meet the needs of digital visual quality assessment of industrial product modeling.