This paper proposes two new measures of three-dimensional shape signatures based on the slope and aspect of three-dimensional triangles, and evaluates the performance of three-dimensional shape signatures derived from distance, area, angle, volume, slope, and aspect for assessing post-earthquake building damage using simulated building models with flat, pent, gable and hip roofs. Three scenarios of post-earthquake building damage are tested using simulated light detection and ranging (LiDAR) points generated on the building roofs. Dissimilarity, sensitivity, and computational cost of the three-dimensional shape signatures are also discussed. The results show that a combination of three-dimensional shape signatures derived from slope and aspect can provide better detection of post-earthquake building damage compared with those derived from distance, area, angle, and volume. The paper demonstrates a promising method for rapid assessment of post-earthquake building damage using existing three-dimensional urban models and post-earthquake LiDAR data.
Proprioceptive shape signatures for object manipulation and recognition purposes in a robotic hand
