Abstract: A feasibility study of the detection of metallic pollutants in soil with thermographic measurement techniques is presented in this paper. This study proposes an alternative method to current techniques for detection and identification of contaminated soils by non-destructive testing to reduce costs and the required execution time. For this purpose, step-heating thermography is used as measurement technique. Taking into account the soil thermal models, different pre-processing methods are applied to the captured thermogram sequences to characterize the soil thermal response data; and Artificial Neural Networks (ANN) are used as a processing tool to discern the presence or absence of contaminants in soil. The selected ANN configuration will determine the contaminated soil identification rates, making the false negative rate worse with the false positive improvement.

Authorship

Download reference