摘要
鉴于利用彩色视觉装置进行在线测量是颜色测量在工业应用发展的趋势,而光源的光谱分布和相机响应函数与人眼的光谱响应曲线的不一致会严重影响测量的精确性,在人工D65光源和LED阵列光源照明下,利用相同的工业相机获取校正图片,采用多项式回归方法在与设备无关空间sRGB和CIEL*a*b*空间进行校正。实验结果显示在校正前平均色差分别为27.68ΔE和16.09ΔE,人工D65光源照明下获取的图片色差较小,校正后不同光源下获取图片的色度趋于一致,分别为2.56ΔE和2.39ΔE。采用sRGB颜色空间,在校正精度、校正步骤、图片显示等方面都要优于CIEL*a*b*空间。以一幅拍摄的实际图片校正为例,显示了色度校正对彩色装置的重要性。
Automatic vision measurement by color vision device is the development trend of color measurement for industrial application. The average color reproduction error of illumination and the spectral response functions of camera have a strong impact on the measurement accuracy. Therefore, the correction images were obtained by the same camera under the artificial D65 illuminator and the LED array illuminator, respectively, the color correction was performed by polynomial regression method in standard red green blue (sRGB) and CIEL ~ a ~ b ~ device-in- dependent color space simultaneously. Results showed that before correction, the color differ- ences of images were 27.68AE and 16.09AE, respectively, meaning illuminated by artificial D65 illumination was less than illuminated by LED array illumination; After correction, the color of images obtained by different illuminators was close to each other, the average color differences were 2.56AE and 2.39AE, respectively. Using sRGB as the device-independent col- or space was better than CIEL ~ a ~ b ~ color space due to the high accuracy, the simple correction process and the convenience of image display. Finally an actual picture captured by color measurement device was corrected in sRGB color space. The results showed the importance of correction for color vision device.
出处
《应用光学》
CAS
CSCD
北大核心
2013年第6期968-973,共6页
Journal of Applied Optics
基金
国家自然科学基金(51305137)
江西省自然科学基金(20122BAB216025
2010GQS0032)
江西省教育厅青年科学基金(GJJ12316
GJJ13320)
关键词
颜色校正
颜色空间
彩色视觉
color correction
color space
color vision