Person Identification by Evaluating Gait using 2D LiDAR and Deep Neural Network

Abstract

Video-based recognition techniques are solemnly effective, and it comes to a new era of research nowadays. Yet again, it suffers some bottlenecks indeed. Situations, surroundings, and momentums may be disgraceful with all new inventions. So, to solve the drawbacks of technology is to imply a new technology on it. Biometric features are very authentic and high valued measures for human identifications. Most of the techniques are dependent on close contact with the subject. A gait is a pattern that performs by walking from the individual. Almost all studies of gait-based person identifications are performed by RGB or RGB-D cameras. Very few studies were done by using LiDAR data. Applying 2D LiDAR images for individual tracking and identification is superb when video surveillances fail to perform accurately due to environmental and imposed difficulties (i.e., disaster, rain, fog, smoke, snow, occlusion, cost, etc.). This research performed a comprehensive exhibition of 2D LiDAR data with a rigorous self-made dataset and customized residual neural network. We considered different experimental setups and found exciting precisions there. Our system is appropriate for recognizing a person based on his ankle level 2D LiDAR data.

Single Person Identification and Activity Estimation in a Room from Waist-Level Contours Captured by 2D Light Detection and Ranging

“…The alternative method involved generating an image from a point cloud and utilizing a CNN-based image classification approach for the generated image. The image generation method followed a similar approach to gridmaps used in previous studies [ 11 , 22 ]. Specifically, scan points were projected onto a map with a 0.01 m grid on each side.…”

“…In a previous study using 2D-LIDAR at ankle height, 92.4% accuracy was achieved for three-person identification [ 18 ]. Another study realized 98% accuracy in 29 person identification using multiple 2D-LIDARs at ankle height [ 11 ]. In comparison, the accuracy of person identification in our methods was slightly lower.…”

“…Notably, LIDAR can extract sufficient features from contour point clouds [ 10 ]. Person identification has been realized using both 2D-LIDAR [ 11 ] and 3D-LIDAR [ 12 , 13 , 14 ]. Combining 2D LIDAR and cameras has also proven effective for person identification [ 15 ].…”

“…Although 2D-LIDAR captures less data for person identification and action estimation compared to 3D-LIDAR, research using 2D-LIDAR at ankle height has successfully achieved person identification. The sequence of ankle contours reflects gait features adequately [ 11 , 18 ]. Person identification research using 3D-LIDAR also suggests that the lower part of the body improves identification performance [ 13 ].…”

“…For instance, leg contours were detected from an occupancy map created from 2D-LIDAR point clouds [ 22 ]. In [ 11 ], an image was superimposed on a point cloud of a person walking, and person identification was performed using a CNN model for the superimposed images.…”