图像RGB数值衍生的欧氏距离用于手机比色法检测石油类物质

Determination of Petroleum Substances by Mobile Phone Colorimetry Based on Euclidean Distance from Image RGB Value

利用手机建立了一套石油类物质浓度的检测系统,将基于RGB值和空白值计算得出的欧氏距离D值应用于图片比色,提高了方法的精密度。分析了不同性能的手机在高照度下的检测效果,结果说明大光圈值(F)手机拍摄得到的图片比色效果更好。考察了不同白平衡模式和光源对D值标准曲线的影响,获得了适宜的条件,即对于偏暖色系的浓度检测,色温2800 K的白平衡模式和300 W碘钨灯有较好的效果。该方法的线性相关系数在适宜条件下大于0.99,测试范围与分光光度法相当;测定标准样品的相对误差为2.13%,相对标准偏差为3.0%,加标回收率为92.0%~105%。该方法准确方便,成本低,易维护,可实现远程自动化的实时在线检测。

In this paper,a method was established for the quantitative detection of petroleum substances using mobile phone as detector.D values for Euclidean distance based on RGB and blank value calculation were applied to colorimetry to improve the precision of the method.Influences of different smart phones on the detection results under high illuminance were investigated.Results showed that the detection effects were significantly improved with the increase of aperture value(F).Influences of white balance mode and light source on the standard curve of D values were investigated to obtain the suitable conditions for warm color system detection,namely a 2800 K white balance mode and a 300 W iodine tungsten lamp as light source.Under the above conditions,this method was applied in the detection of standard sample with a correlation coefficient larger than 0.99 and a testing range similar to that of spectrophotometry.The relative error for standard sample was 2.13%and the relative standard deviation was 3.0%,and the recoveries for standard addition ranged from 92.0%to 105%.This proposed method is accurate,convenient,low cost and easy to maintain,and it is suitable for real-time online detection of remote automation.