Error reduction in arsenic detection through color spectrum analysis

Groundwater contamination by Arsenic is a huge problem in many countries. In a developing country like Bangladesh, with widespread Arsenic contamination and lack of laboratory facilities, usually field detection kits are preferred to detect arsenic in tubewells. However, these kits produce a significant number of false positives/negatives due to human errors in matching the detection test-strip colors to the reference color chart. This paper introduces digital image processing methods and a smartphone application, which allow fast and inexpensive improvement in the test-strip classification of field detection kits. A smartphone captures a photo of the test strip used in the field detection kit, while the application detects the Arsenic level by comparison with reference colors. This automation reduces human errors while matching the colors using the eyes only, by adding an extra layer of cross-checking. Thus, the overall accuracy of the Arsenic detection process is improved.