有色示踪剂模拟槽实验图像分析法

Image Processing Method of Dye Tracer Experiment in Sand Box

图像分析法是处理有色示踪剂模拟槽实验结果的常用方法之一。本文以食用亮蓝为示踪剂,从污染物的检测精度和污染羽描绘的准确度两个方面将图像分析法与传统取样法进行对比,并利用Matlab编程对模拟槽内的污染物贮存质量进行定量计算。结果表明,与传统取样法结果相比:图像分析法的检测精度可提高1个数量级;污染羽的描绘准确度至少可提高3%左右,在实验所设置的非均质条件下可提高17%左右。利用自行编制的Matlab程序对本实验模拟槽内贮存亮蓝质量进行计算可得,当剖分格数为3 7002(1 369万)时,亮蓝贮存质量的计算值最接近实测值,最小误差仅为0.5%。综合来讲,图像分析法是一种准确度高、空间分辨率高的信息采集和处理手段,可实现对非均质地层中有色示踪剂运移过程更为精细的描述。

Image processing method is commonly used in analyzing the experimental data obtained from dye tracer experiments.The authors contrasted the image processing method with a conventional sampling method to monitor the precision and plume accuracy in using brilliant blue as the tracer.We quantitatively calculated the mass storage in the sand boxes by the Matlab programs.The results showed that the monitoring precision and the plume accuracy raised by one order of magnitude and about 3%respectively when we used the image processing method,and the plume accuracy raised by 17% under the heterogeneous conditions set by the experiment.The storage mass was almost the same as the one using the Matlab program.When the grid number was 13.69 million,the minimum error was 0.5%.Generally,the image processing method provides a high-accuracy and high-resolution way to acquire and dispose the picture to describe the migration process in the heterogeneous media more subtly.