Wiedemeijer, M.M. (Maria) (2019) Evaluation of Expert Observations of Gait in Forensics. thesis, Human Movement Sciences.
Full text available on request.Abstract
Introduction Forensic Gait Analysis (FGA) is a growing topic in forensic science. Since computer vision does not su�ce, the expert eye is required. To be acceptable as proof in court it is necessary to know how good observers are able to score features of gait. Methods A computer application was developed: GOMI (Gait Observers Measurement Instrument), o�ering animations of a walking avatar with varying features of gait, which are elements of a gait pattern. This application was tested on a group of observers, o�ering 13 features in 5 values in 3 camera positions. With a logistic regression model score errors were modelled as a function of score session, gait feature, feature value and camera position. Results Observers show a wide range in gait feature scoring capability from random to practically maximal. Camera position is of importance for the scoring of features, consistent with anatomical planes. Gait features di�er in scoring success by the observers by a factor of maximally 10. Observers and gait features appear to largely determine score errors. Conclusions A tool has been developed that can distinguish experts on their scoring capability of features of gait and shows a dependency on camera position and type of gait feature in that capability. These results help to progress the �eld of forensic gait analysis. Keywords: forensic science, walking, gait, forensic gait analysis, gait recognition
| Item Type: | Thesis (UNSPECIFIED) |
|---|---|
| Faculty: | Medical Sciences |
| Date Deposited: | 20 May 2022 11:42 |
| Last Modified: | 20 May 2022 11:42 |
| URI: | https://umcg.studenttheses.ub.rug.nl/id/eprint/3389 |
Actions (login required)
 |
View Item |