South Wales Police to scan EVERY fan's face at Champions League final next month

Police to deploy real-time facial recognition against football fans at Champions League final in June

South Wales Police is planning to scan every face in the audience at June's UEFA Champions League final as part of its security operation.

The Force will use facial recognition technology to compare the faces against their database of around half-a-million "persons of interest".

With no British sides left in the Champions League after Leicester City was knocked out by Atletico Madrid at the quarter-final stage, the operation is likely to be geared more towards a potential terrorist threat than everyday crime.

Fans' faces will be scanned both at the stadium and at Cardiff Central railway station on the day.

According to a £177,000 tender for an automated facial recognition solution, which closed in March, the contract will last for two years, from April 2017 to April 2019, but is explicitly intended to be used at the Principality Stadium on 3 June in a pilot project.

"Programme Fusion is the collaborative technological programme between South Wales Police and Gwent Police," reads the tender.

It continues: "As part of the programme it is the intention to effectively utilise the advancements in automated facial recognition (AFR) technology.

"The strategy for AFR falls into two broad areas. Real-time facial recognition and slow-time static face search. Real-time facial recognition being specifically linked to a live pilot on Saturday 3rd June 2017 in and around the principality stadium and Cardiff central train station on the day of the UCL Champions League Final.

"The slow-time static face search being specifically linked to the 500,000 custody images stored within the Force's Niche Record Management system."

The plans and the tender were uncovered by the Motherboard tech website.

However, South Wales Police will need to abide by the Surveillance Camera Code of Practice, published in June 2013, and there are question marks over whether a project that has been put together so quickly will be able to provide accurate facial recognition.

"Getting the best, most accurate results for each intended application requires good algorithms, a dedicated design effort, a multidisciplinary team of experts, limited-size image databases, and field tests to properly calibrate and optimize the technology," the UK government surveillance camera commissioner Tony Porter told Motherboard.

On top of that, the accuracy of facial recognition remains open to question, especially in circumstances where the subjects are part of a crowd, rather than standing face-to-face staring into the camera.