One of the things that fascinate me the most in forensic video analysis is the relation between the subjective digital data and the objective human interpretation involved in any investigation. Psychological biases and the fallacies of human perceptions easily verifiable with any of the popular optical illusions are just some of the factors which must be taken into account while doing investigations.
But this time I want to look at things from a higher level and talk about the usefulness of video as evidence and our perception of it. Chances are you have already seen the very interesting article: “The Value of CCTV Surveillance Cameras as an Investigative Tool: An Empirical Analysis” (link).
The abstract provides some impressive numbers: “This study sought to establish how often CCTV provides useful evidence and how this is affected by circumstances, analysing 251,195 crimes recorded by British Transport Police that occurred on the British railway network between 2011 and 2015. CCTV was available to investigators in 45% of cases and judged to be useful in 29% (65% of cases in which it was available).”
For reference, this is the decision workflow used in the classification (image from the above paper).
This really made me feel good. It looks like what we are doing here at Amped Software is having an impact on society, and more than we expected. I think most people in our community would be surprised by the numbers. At Amped, we see hundreds of cases every year, and for more than half of the images and videos that we receive, we just say that they are useless.
I often like to say that what attracts people the most to Amped FIVE are the amazing results of our sample images but what people appreciate most once they start using Amped FIVE on a regular basis, is the seamless and scientific workflow for everyday tasks. Most of the material we get is the classic three-pixels-saturated-compressed-blurred-license-plate, and there’s not much to do there anyways if you just want the license plate.
So, we often complain about the quality of CCTV footage, the inexperience of the system installer or the decision of the system owner to buy cheap stuff. We don’t realize, even in the bad cases, how much information we can actually get from CCTV, even if the quality is terrible.
Below are some additional figures from the paper which I think are worth noting.
In the first image below, you can see that the quality was not the only issue, but also technical problems like overwriting and a faulty system are a big deal, and many times CCTV was not requested at all.
In the next image below, you can see the impact of CCTV on several types of crimes. There’s a long discussion on various types of crimes in the paper so I highly recommend to read it thoroughly.
Citing again the paper above: “A criminal investigation can be thought of as a series of questions: who was involved in an incident, where did it happen, what happened, when did it happen, why did it happen and how were any offences committed, known as the ‘5WH’ investigation model (Cook et al. 2016; Stelfox 2009).”
Very often the quality of a video is insufficient for identifying a person or a license plate, but the video can give us other important information: if some event is filmed, even at very low quality, you can almost always get an idea of the what, when and where. Sometimes also the why and how can be inferred by the video (think about a car crash or a fight in a bar) even if the quality is not that good. Also, missing information could be good information… at least we can exclude a location identified as suspect, if there is no relevant footage from the camera in that location. And often, even if the quality is not good for identification, we can still get a lot of information from the video: that he was a man and not a woman, that it was a black van and not a white car and so on. We should be more positive and see the glass half full and not half empty… actually I think it’s full for more than three-fourths… probably.
There is another interesting note in the paper. In the past, there have been other studies which show very little impact from CCTV on investigations, however the authors justify it as follows, and it totally makes sense to me: “it appears that the apparent low usefulness of CCTV reported in previous studies such as the unpublished report produced by the MPS may be a function of CCTV being only infrequently available to some investigators. The present study was able to distinguish between the availability and usefulness of CCTV, in contrast to other work that has not made that distinction, artificially lowering the apparent usefulness of camera systems. Failing to make this distinction is somewhat akin to saying that witness evidence is not helpful in cases where no witnesses were present: true, but not particularly insightful.”
And, to conclude, I found this last piece a bit worrisome: “However, it is possible that there remain cases where CCTV could have been useful but in which the investigating officer did not make a request for it. A recent small-scale study of newly installed cameras in two violent-crime hotspots in Stockholm found that investigators requested CCTV recordings in only 20% of cases in which they were available (Marklund and Holmberg 2015). It is crucial that officers take reasonable and proportionate steps to identify CCTV evidence in every case in which it may be available.”
Clearly, analyzing CCTV is costly and time-consuming. With our tools, we are trying to minimize the efforts required, but it’s a pity that potentially useful evidence is not getting considered at all. Of course, probably doing an in-depth investigation for a guy who stole a bike costs way more than the bike itself. But what are the chances that this guy will steal other bikes, then cars and maybe someday a robbery? Some food for thought…