TDS Mobile Vehicle Identity Recognition System
The TDS vehicle identification system combines advanced license plate reading and make/model determination technologies to provide a very powerful tool for identifying vehicles of interest from a mobile platform such as a police car.
The Vehicle Identity Recognition system consists of a police car roof-top mounted camera system that is interfaced with a trunk located computer processor. The system includes a GPS and a 4G LTE modem allowing communications with remote servers for cloud processing, blacklist/other updates and event notification as required. The system is designed to operate in a fully automated mode and will continuously acquire and process vehicle images captured from the roof mounted camera. The images will be processed using TDS provided automated license plate recognition (ALPR) software which will accurately derive the license plate number. The system compares the resultant license numbers to determine when a match occurs against a pre-established blacklist. When a plate match is found, the system additionally determines the make and model of the vehicle (VMM System) as based on the image(s) acquired for that vehicle. The license plate, make/model and blacklist entry information is used to generate a notification message that is transmitted by WiFi for display as an alert on the police officer tablet display. The matching event information can additionally be forwarded to a remote server and may include the GPS and time stamp. The location of the police car can be updated periodically to a central vehicle location system where a location map of all police cars in the system can be generated and displayed for command and control purposes. The trunk located processor runs the Linux operating system and can run additional police related applications. The police officer's tablet display has complete access to the internet via the 4G link and can browse the internet and specific intranet application servers
The basic concept of the VMM solution is to match a query image to one or more predefined set of classified images that have been indexed and trained to be recognized by the detection platform. If there is a match to any of the sample images, then the application confirms with a certain level of confidence that the query image is of the same class as the matched image. The Vehicle Make and Model solution is provided as a service through an HTTP server that performs the query on a cluster of computers working together on each request to deliver the classification results in a less than a second. The service will be installed on a cloud computing provider (e.g. Amazon AWS). The request to detect the make and model of the vehicle returns the following attributes:
- Make - The vehicle make (e.g. Ford, Toyota)
- Model - The Vehicle Model (e.g. F150, Camry)
- Year Range - The year range in which the shape of the vehicle model from a specific view is the same.