An important article over at Australia's ABC Science, reporting on the work of Professor Gary Edmond from the University of New South Wales. Professor Edmond is the author of an article in the latest edition of Current Issues in Criminal Justice in which significant questions are raised about the all-too-common court use of video and photographic footage from CCTV or ATMs as evidence in criminal cases.
The Professor believes that images compared to those of a prime suspect are carried out in a way that may encourage prejudice in favour of the police view:
"The courts don't impose a reliability standard on the use of these images as evidence," Professor Edmond said.
As Professor Edmond points out, images are usually compared to those of a prime suspect, but are also sometimes compared with photos of a range of people to eliminate suspects – on the basis of body height, for example. "Experts" in anthropology, anatomy and photography carry out the comparisons and are sometimes referred to as "face and body mappers" or "photo comparison experts". Juries also draw on evidence of images sometimes after listening to expert interpretations of these.
But Professor Edmond says there are two problems with the use of image comparison:
1) the procedures used for comparing images have never been validated, and comparisons are carried out in a way that may encourage prejudice in favour of the police view.
Professor Edmond says there have been cases where image comparison by British experts have been found to be inaccurate, that reality is distorted in images.
It may be reliable when a suspect has a distinguishing feature and there are good quality CCTV images, but when images are poor and experts are comparing facial features alone, there are serious complications.
The shape and size of someone's nose, for example, can look different depending on the distance they are from the camera, the angle of the camera, the type of lens used, shadows and lighting.
"It's not as straight forward as people often claim," Professor Edmond said. "Similarity in images may not be similarities in reality."
Professor Edmond says photo comparison experts have never provided any evidence to support the validity of their conclusions.
He says that courts should require tests that show how reliable experts are at identifying people by comparing images.
2) The police often tell photo comparison experts who their prime suspect is, compounding the psychological tendency to find similarities when given just two photos to compare.
Professor Edmond believed that in order to reduce such bias, it may be useful to use two experts to make a comparison. One expert could analyse the features of a CCTV photo, without seeing the prime suspect's photo, and their analysis could be compared with that from another expert viewing a photo of the prime suspect to see if there was any overlap.
The Professor believes that judges are reluctant to exclude potentially valuable evidence from a trial and the increasing availability of CCTV is providing more and more potential image evidence.
"The increasing availability of (so-called) evidence means the courts are responding regardless of whether they've actually got a credible way of dealing with it," he said.
"I think the technology is driving the law, even though the question of its reliability may not be adequately resolved yet."
Potential bias could be dealt with by both the prosecution and defence having their own photo comparison experts. But that would unfairly put the onus on the defendant to prove the method is unreliable.
"And it assumes the existence of a credible field when that is yet to be demonstrated."
Combined with all the other doubts about CCTV we've raised, it's food for thought, isn't it?
By Alex Deane