Leslie Bowie, Stephanie Plews and Michael Bromby weigh up the pros and cons of presenting CCTV footage as evidence

The UK has the highest concentration of CCTV cameras in the world and there is seldom a major crime on the streets of our cities that is not caught on video.

In Britain, there is now an estimated one camera for every 15 people, which amounts to more than four million in total.

Today, there are few cases that do not involve CCTV evidence, but how can it be used to maximum effect? As evidence in court, a CCTV image can often be difficult to decipher.

As well as changes in clothing and hair that are an easy way for assailants to disguise themselves, the effects of camera lenses, street lighting and shadows can often make it difficult for a jury to judge the exact identity of a suspect in a CCTV image.

To an untrained eye, these nuances can often be undeterminable.

The majority of identification work from CCTV is the comparison of a perpetrator caught on camera, with a suspect.

This is commonly termed 'facial mapping', and is a practice employed by around 12 practitioner groups within the UK.

Methodology differs among practitioners, but in general, an assessment of the facial geometry and morphology is undertaken and a conclusion reached as to the likelihood that the two images being compared are in fact those of the same person.

With high quality CCTV footage, there will be little dispute as to whether a suspect was involved in an incident, and this type of imagery does not require interpretation by an expert, as outlined in the appeal cases of R v Dodson and Williams [1984] 79 Cr.

App.

R.

220.

However, with poor quality facial imaging, there is a need for facial imaging expertise.

Poor quality footage can be scientifically analysed, permitting the CCTV image to be used as an investigation aid to identify a suspect, or as an admissible source of identification evidence in court.

Poor quality imagery may be a result of the over-use of videocassettes, low-resolution camera systems, the proximity of the subject or incorrect lighting levels.

The expert can interpret the footage in light of these variables by employing techniques such as image enhancement and zoom effects.

In addition, motion-blur, changes in facial expression and head orientation are factors that are taken into account in assessing the imagery for identification purposes.

An expert's conclusion may be based on a number of factors.

These include: the quality of the imagery (how much facial detail can be seen); the distinctiveness of certain facial features and the prevalence of the features within the population; the number of matching facial features that can be seen (in particular the combination of these observed features); matching facial geometry and the number of matching viewpoints that can be observed (how many different views of the face can be matched so that a pseudo three-dimensional picture of the person is constructed).

Much of the work undertaken by practitioners involves facial identification.

However, there may be significant background information that can be enhanced or studied.

Objects, clothing, vehicles or indeed animals, may be compared by applying similar imagery analysis techniques.

Experts may also be able to calculate an individual's height or distance from a given point from an image.

Practitioners have been presenting evidence from facial imagery analysis for around 15 years in all three jurisdictions of the UK.

In addition, the Appeal Court has considered it in evidence on a number of occasions.

The earlier cases of R v Ryan (Unreported, 1991) and R v Stockwell [1993] 97 Cr.

App.

R.

260 established the admissibility of 'facial mapping'.

The Attorney-General's reference (no.

2 of 2002) identified four competent means of establishing identity from CCTV footage.

These are:

- Jury members may make their own comparisons of crime scene images with the defendant without the need for additional expertise (R v Dodson and Williams, 1984).

This often occurs with high quality imagery.

- A witness known to the defendant may recognise that defendant from the images and give evidence accordingly (Taylor v Chief Constable of Cheshire [1987] 1 All E.R.

225).

- A witness not known to the defendant may make repeated viewings of the imagery and acquire special knowledge of the footage that can be presented to the court (R v Clare and Peach [1995] Cr.

App.

R.

333).

- A suitably qualified expert with facial mapping skills can give opinion evidence of identification (R v Stockwell, 1993).

Recently, R v Gray (Court of Appeal, No.

2002/2459/X3, 27 March 2003) questioned the objectivity of facial mapping.

It was noted that a 'national database of facial characteristics or any accepted mathematical formula from which conclusions as to the probability of the occurrence of particular facial characteristics or combinations of characteristics can be drawn' should be employed by an expert when considering identity.

Such expert testimony was regarded as 'both admissible and frequently of value...

to permit the jury to reach its own conclusion'.

Commissioning imagery analysis often leads to a change of plea because of the damning nature of the evidence presented in a pictorial format.

However, when presented in court, imagery analysis is generally used in combination with other corroborative sources of identification.

In R v Mitchell (unreported, Portsmouth Crown Court, 13 Aug 2003), the expert's opinion was sufficient for the jury to convict on that evidence alone.

This was also observed and upheld by the Court of Appeal in R v Hookway [1999] Crim.

L.

R.

750.The objectivity of imagery analysis can be upheld by the adoption and maintenance of accepted practices within the field.

In 2003, the Forensic Imagery Analysis Group (FIAG) was established under the auspices of the British Association for Human Identification.

This is a self-regulatory group of practitioners aiming to validate their practices with the Council for the Registration of Forensic Practitioners.

The group was instrumental in the revision of the Association of Chief Police Officers national working practices in facial imaging.

FIAG seeks to discuss the limitations and possibilities of imagery analysis.

Currently, there is agreement that positive identification cannot be made unless unique features such as marks and scars or tattoos are observable.

Proof of non-identification can only be made when differences are observed that cannot be explained by extraneous variables, such as lighting or viewpoint.

Between the two ends of the spectrum, levels of support can be given for the hypothesis that the two images are of the same person.

The National Crime and Operations Faculty maintains a database of practitioners, although this does not formally qualify or recognise persons or practices.

Many practitioners are also listed in various expert witness directories.

Leslie Bowie and Stephanie Plews, are experts at ABM-UK's facial verification bureau.

Michael Bromby is research fellow at Glasgow Caledonian University in the Centre for Forensic Statistics and Legal Reasoning and is group secretary of the Forensic Imagery Analysts Group