轮辋类螺纹表面成像的光照方法研究

Illumination method of rim thread-like surface imaging

针对摩托车车轮结构多样性导致外观质量检测的自适应成像空缺的光照问题,提出了一种基于类螺纹剖线灰度分布的单位牙幅值评估方法来优化轮辋光照方向,能够采集到清晰的加工纹理图像。首先,对轮辋的加工纹理表面剖线建立了以牙为单位的类螺纹剖线几何模型;其次,结合单位牙结构和轮辋结构的阴影条件约束了光照方向,并在类螺纹剖线的几何模型上模拟了不同方向的光线产生的光亮度;最后,根据建立的亮度差异公式获取类螺纹剖线的灰度分布中幅值最大化的光照方向。对提出方法进行有效性评估实验,结果表明不同方向的光照对单位牙产生不同程度的亮度差异;由于秩和分布的差异统计量Zstatistic<1.96,表明所提的亮度差异公式符合实际;光照方向优化实验表明轮辋在-120°光照方向下类螺纹纹理及其他缺陷均能清晰成像。

The illumination problem of self-adaptive image vacancy in appearance quality detection is mainly caused by the diversity of motorcycle wheel structure. To solve this problem, an evaluation method of unit thread amplitude based on the gray distribution of rim thread-like profile line is proposed, which can optimize the illumination direction and obtain clear appearance processing texture images. Firstly, the geometric model of the thread-like profile line based on thread is formulated. Secondly, in the light direction interval determined by the unit thread and the rim structure, the illumination brightness in each direction is simulated on the geometric model of the thread-like profile line. Finally, according to the established luminance difference formula, the optimal illumination direction that maximizes the amplitude in the gray distribution of the profile line is achieved. Experiments are implemented to evaluate the effectiveness of the proposed method. Results show that different directions of light result in different levels of brightness difference of unit thread. Because of the difference statistics zstatistic<1.96 of rank-sum distribution, results show that the luminance difference formula proposed in this paper conforms to the reality. The light direction optimization experiment shows that the rim with the thread-like texture and other defects can be clearly imitated under the-120° light direction.