Phagocytosis-based camera-in-situ method for pile load testing

• A vision-based approach is presented to record the pile-head settlement under ASTM’s standard.
• An immune-inspired algorithm is designed to search the nodes in successive-sparse images.
• A detailed procedure to conduct the in-situ test is also presented.
• A detailed comparison is done to validate the proposed methodology w.r.t. LVDT-based approach.

This paper reports a novel camera-in-situ methodology to calculate pile settlements through nodal displacements, using a biologically inspired mechanism of phagocytosis to correlate successive images. An x–y grid is placed on the pile-head and high resolution images are captured through a DSLR camera, remotely. Software is also developed and presented herein which automatically interprets the captured-images into nodal pixel indices and subsequently into actual nodal displacements. The software adopts a methodology analogous to that of a monocyte searching for bacteria in its vicinity, to search each node in successive images. Moreover, conventional linear-variable-differential-transformer-based methodology is adopted to validate the results of the presented technique. It is found that the presented approach is in conformance with the guidelines set forth by ASTM standard for static pile load testing, in addition to the advantages of being completely autonomous and free of any embedded instrumentation on the pile.