Non-Contact Sensing System to Measure Specimen Volume During Shrinkage Test

The shrinkage curve provides information of value for the interpretation of soil-water characteristic curve data. However, there is need for an accurate and precise volume measurement technique during the shrinkage test. This paper presented an inexpensive automated digital image processing technique, which allowed accurate and precise measurements of the soil specimen volume. The volume was computed by accurately measuring the radius and height over the entire lateral surface of specimen. The proposed volume measurement technique involved projecting a structured light laser on the soil specimen in order to provide the reference points for measurements. A 360° view of the specimen was then captured using a camera. High resolution images were then processed using functions developed within MATLAB. The computations also allowed the reconstruction of a 3D mesh model of the specimen. A validation test on a dummy object showed that the error in radius and height measurements at more than 75 % measurement points was less than ±0.05 and ±0.07 mm, respectively. A 99 % accuracy was achieved in the volume measurement of the soil specimen.