Forensic podiatry is a medical discipline that plays a crucial role in the identification process in legal proceedings requiring specialised knowledge in this field [1]. Crime scenes often present evidence that needs to be evaluated from a forensic podiatry perspective, such as plantar footprints (partial or complete), footwear, anatomical remains of the feet, biomechanical analyses, or images and recordings of gait captured by Closed Circuit Television (CCTV) cameras, all of which can contribute to clarifying the facts [2, 3]. Forensic podiatrists have the expertise to identify evidence related to the feet and assist in identifying individuals through their anatomical features [4, 5]. The principles and knowledge of forensic podiatry are highly beneficial for investigators at crime scenes, particularly when dealing with evidence related to feet, footwear, and/or gait. Moreover, this knowledge helps establish the physical or biological profile of a suspect for individualisation and identification [1, 3].

Forensic gait analysis has emerged as a recent specialisation within forensic podiatry, becoming increasingly recognised and utilised by experts [6]. This analysis is carried out through a structured method, which includes the phases of analysis, comparison, evaluation, verification, and the subsequent preparation of reports [3, 4]. In CCTV gait studies, the gait of a subject and its inherent characteristics are analysed and compared to contribute to the identification process [7, 8]. It is argued that each person exhibits a repetitive and distinctive gait pattern, which gives it an individualistic and recognisable quality. This pattern can allow a person to be identified from a distance by their walking style, making gait analysis a valuable tool in forensic investigations, and emerging as a biometric analysis parameter [9]. In addition to gait, footprints recovered from the crime scene, such as bare footprints, shoe prints, or sequences of footprints, also serve as critical evidence [1, 3, 10].

The use of CCTV cameras and other surveillance methods has generated growing interest in the practice and research of forensic gait analysis and its potential for human individualisation. For the first time, forensic gait analysis was used as admissible scientific evidence at the Old Bailey Central Criminal Court in London, UK, in the case of R v. Saunders in 2000 by the UK-based forensic podiatrist, Dr. Haydn Kelly [6, 8]. Since then, various countries have admitted gait analysis as a complementary form of evidence in court, although this method is still considered questionable in terms of reliability and accuracy [3, 6, 8, 11]. Similarly, gait analysis has been used in forensic practice in many countries, and camera-based computer systems have been developed to recognise a person’s highly unique movement patterns, offering distinctiveness to their gait [12]. Currently, gait recognition systems based on artificial intelligence, such as convolutional neural networks, are being implemented and have been shown to improve identification accuracy, even with low-resolution recordings [13].

Today, the majority of crimes occurring on public streets are highly likely to be captured by video surveillance cameras located in shops, banks, or other public places. CCTV images provide different views of the recorded subject (front, transverse, and sagittal, with the frontal view being the most reliable [14]), eliminating the psychological factor that might predispose an individual to simulate an unreal gait, as individuals are usually unaware they are being recorded. As a result, this method is non-intrusive and does not require the collaboration of the subject, making it particularly useful in the field of criminalistics [14, 15]. Before the invention and use of CCTV, investigative agencies relied primarily on eyewitnesses who claimed to have “seen” someone at the scene with a particular “walking style.”

In forensic gait analysis, angular measurements play a key role, as they are necessary for the analysis and the subsequent presentation of reports as evidence in legal proceedings [3, 16, 17]. Examining these angular measurements of the gait cycle is essential, as they provide quantitative and objective information about the study performed. The joint angles and those formed by various body angles have an individualistic character for each subject [18]. For an exhaustive analysis of video evidence, it is necessary to use specific software that allows tracking key points of the body and calculating joint angles during the gait cycle, such as those of the hip, knee, or ankle [18,19,20]. When video footage only shows a partial gait cycle (since it represents only a portion of the full cycle), evidential recognition using the observational method becomes challenging. In these cases, the use of joint measurements at specific time intervals is advocated, aiming to study the variability between suspects and obtain discriminatory parameters [21].

Therefore, gait analysis has significant potential to provide valuable evidence, as it is a unique characteristic of each individual [20]. However, improving this analysis with accurate measurements is complicated due to the low quality of some surveillance footage, which prevents a conclusive identification of a person based solely on image analysis. It cannot be definitively stated in a court of law that no other person could have a similar gait pattern based on a specific set of characteristics [3, 22]. As a result, gait analysis is not considered as strong evidence as fingerprints or DNA, but it can be useful in the absence of conclusive evidence. However, when combined with photogrammetry, gait analysis can be regarded with greater reliability [23]. Therefore, the aim of our study is to evaluate the reliability of forensic gait analysis by using angular measurements for subject identification and to compare it with observational analysis of morphological features.