Exploring the relationship between stride, stature and hand size for forensic assessment

Area of Science:

• Forensic Science
• Biometrics
• Anthropometry

Background:

• Forensic evidence frequently relies on accurate measurements, estimations from landmark data (e.g., stature from image data), and inferred data.
• Surveillance videos often capture limited physical evidence, necessitating methods to infer key measurements from available data.
• Automated gait analysis systems can extract stride data, providing a potential link to other physical characteristics.

Purpose of the Study:

• To explore the correlations between hand measurements, stature, leg length, and stride length using a novel dataset.
• To develop predictive models for estimating stature from hand size, leg length, and stride, and vice versa.
• To assess the impact of a subject's sex on the accuracy of these predictive models.

Main Methods:

• Utilized a novel dataset linking hand measurements, stature, leg length, and stride.
• Employed correlation and regression modeling to establish relationships between variables.
• Incorporated stride data extracted from an automated gait analysis system.
• Evaluated model performance with and without stature as an intermediary variable, and with known subject sex.

Main Results:

• Established accurate predictions of stature from hand size, leg length, and stride length, and vice versa.
• Demonstrated the ability to predict leg and stride length from hand size, with or without stature as an intermediary.
• Showed improved prediction accuracy when the subject's sex was known a priori.
• Validated the potential for calculating or verifying relationships between suspect physical measurements from limited data.

Conclusions:

• Developed robust models for inferring physical measurements in forensic contexts, particularly from surveillance footage.
• Highlighted the utility of integrating hand measurements, leg length, and stride analysis for biometric estimations.
• Confirmed the value of incorporating demographic data, such as sex, to enhance the accuracy of forensic predictions.