A relatively new field of forensic science in India, gait analysis helps identify a suspect by comparing their walking pattern with CCTV or other video footage from a crime scene.
But this method has hardly stood the judicial test of evidentiary value. This is how gait analysis works, and why it is considered corroborative rather than standalone evidence.
Gait analysis is a sub-division of the field of forensic science known as “forensic podiatry”. This is the school to which bare footprint analysis also belongs.
Police in India are increasingly using gait analysis to identify individuals whose identity may be unclear from crime scene video footage, forensic experts say.
“Unlike DNA and fingerprints, gait is a behavioural biometric that involves capturing the overall dynamics and kinematics of an individual’s running or walking pattern, requiring no physical contact for capturing the data,” says Anshu Singh, assistant professor at the School of Behavioural Sciences and Forensic Investigations, Rashtriya Raksha University, Gujarat.
While the evolution of gait analysis dates back to the 1970s in Europe, the first recorded case employing the gait pattern as an admissible evidence was recorded in 2000, when a person attempting to rob a London jewellery shop was identified by this method.
“While by mid-2000s many cases were solved with the help of gait pattern analysis but simultaneously raised questions on its reliability and methodology, also raising questions if manual gait analysis possessed sufficient scientific validity for criminal proceedings,” Singh adds.
How is gait analysed in forensics?
Singh says forensic gait analysis involves two approaches — manual and computer-based. Manual analysis involves analysis carried out by qualified experts like forensic podiatrists or biomechanists who generally compare footage and apply their knowledge of human anatomy and biomechanics to understand the gait patterns.
Computer analysis, on the other hand, employs algorithms to extract, analyse and compare gait patterns — including features such as silhouettes, cadence and stride length — while using deep neural learning networks that are trained on vast databases to look for matches, says Singh.
The database is now expected to expand, with the Criminal Procedure (Identification) Act 2022 and its rules, which replaced the erstwhile Identification of Prisoners Act, permitting collection of “measurements” and retaining such data for up to 75 years, as well as sharing of such “measurements” data across databases.
When and how is gait analysis deployed during probes?
In Delhi, police deployed gait analysis to pin down the accused in the UPSC aspirant’s murder since their faces were not visible in the CCTV footage.
Two experts from the National Forensic Science University, Gandhinagar recreated the exact sequence seen in the footage, including placement of surrounding items such as cars. According to Delhi Police, the test took around 4-5 hours.
“Though being used to solve cases, forensic gait analysis is generally considered as supporting or corroborative evidence rather than standalone evidence (in courts),” says Singh.
“There is a need for more trained and qualified gait analysts, automated gait recognition systems with AI-enabled capabilities to understand and solve such cases. Another add-on can be the collaboration between forensic and computer science departments which can accelerate the progress in the field and provide more powerful identification technologies,” she adds.
A number of high-profile cases in India have used gait pattern analysis to identify the accused.
These include a Special Investigation Team of Karnataka Police ascertaining that a 26-year old activist of the Sri Ram Sena Parshuram Ashok Waghmare being the man who shot at journalist Gauri Lankesh at her home in Bengaluru on September 5, 2017. The analysis was done by the Directorate of Forensic Sciences in Gujarat where the original CCTV footage of the shooting was compared with a recreation of the shooting that was videographed with the suspected shooter following his arrest.
In 2021, in the brutal Sakinaka rape and murder of a woman, the Mumbai police reportedly used the gait analysis test as a forensic tool for the first time. The accused, sentenced to death by a trial court, was zeroed in on by the police assisted by other pieces of evidence, including an eyewitness who established that the accused knew the victim and was seen near the spot at the time.
How have courts in India treated the legality or admissibility of gait analysis so far?
The gait of the same individual can vary significantly, depending on their age and speed of walking, the footwear they are wearing, illnesses a person might be suffering from. The methods too are not standardised, and research papers dealing with gait analysis in the Indian context have cautioned of error rates in its reproducibility and reliability.
For the most part, while dealing with convictions, gait patterns have seen little success at the appellate level of courts.
In 2018, a two-judge bench of the Delhi High Court had overturned a murder conviction where the prosecution’s evidence was largely circumstantial, and among the evidence relied on by the prosecution was a gait pattern analysis of the accused. The court had questioned its reliability, noting that while the court does not possess “requisite expertise to opine on the scientific acceptability, or exactitude, of the manner” in which the analysis was conducted by a private laboratory, “How far such an opinion can be relied upon, to base a finding that the person in CCTV footage was, in fact, the (accused), is highly questionable.”
In June 2024, a Madras High Court while overturning a trial court’s verdict convicting nine accused of murder in 2013, of whom seven were sentenced to death and another two to life imprisonment, had observed “science with regard to gait analysis is not absolute that gait pattern is unique for a person.” The court had also noted that in this specific case, the employee of the private lab where the pattern was analysed, had not compared the gait pattern of the entire body and the analysis was only restricted to the foot pattern.
The Madras HC had also raised questions on the legality of how the gait pattern of the accused was analysed where the prosecution had purportedly compelled the accused to re-enact the scene based on which a demonstration video was taken and then compared with the original CCTV footage from the time of the offence, for gait analysis. The Madras HC had opined that such reenactment, without a magistrate’s order, tantamounts to self-incrimination, and would “amount to giving confession to the police or a confession while in police custody”. The Madras HC had held that thus gait pattern obtained in this way “has no evidentiary value”.
The Madras HC verdict has also been relied on by Karnataka HC as of November 2025.
However, gait pattern analysis is gaining ground, at least in the bail jurisdiction. At least in eight cases, between December 2024 and September 2025 as analysed by the Indian Express, the Calcutta High Court, took gait pattern match with CCTV footage into consideration while deciding bail applications.
(With inputs from Sakshi Chand)
For the analysis, experts compare the CCTV footage from the crime spot with a sample of the suspect’s walk. Wikimedia Commons 
Share
