Effectivness of ALPR in Chicago?
A previous post focused on the extent of Automated License Plate Recognition (ALPR) in Chicago. It noted there were at least 75 locations (both fixed and mobile) operated by the police or OEMC that use ALPR. The post explores the effectiveness of the ALPR technology. For three different sources, here are some statistics on their effectiveness:
In the first ten months of the cameras (starting from January 2006), they had recovered 310 vehicles for the first 2.3 million plates scanned – see Sun-Times from Nov. 2006. By May 2007, there were 725 vehicles recovered and 6.5 million plates scanned. By spring 2009, there were over 13 million plates scanned and 1000 vehicles recovered. (The data also includes arrests). Here is a graph of the recovered vehicles over time:
I was a bit surprised when I saw these results (based on three data points). I would have expected a high rate of recovered cars initially and then a gradual taper to a plateau. First, I don’t have the most reliable data sources. Second, I don’t know much about the circumstances of how these cameras were deployed (how the cameras come online, how they were deployed, where they were use).
The average is one recovered car per day (using over 75 cameras). There are also 325,000 license plates scanned every month on average. If every camera is working equally, this works out to 144 license plates scanned per day with each of the 75 cameras. This is a very low number, because some of these cameras are capable of scanning 3,600 license plates per hour! I have no idea why this discrepancy exists.
I can’t fully explain this data, but I thought it would be useful to publish it. Please let me know if you have any explanations.
Recent Comments