A plain guide to ANPR and ALPR licence plate recognition: how it reads plates, where to place cameras, hot lists, plate journeys and running it inside your VMS.
ANPR (automatic number plate recognition, called ALPR in North America) does four things in a fraction of a second. It finds the vehicle and its plate in the picture, crops the plate, reads the characters, and checks the result against your lists of known plates.
The best systems do not trust a single frame. They follow each plate as the vehicle moves and combine several readings into one confident answer. Five frames that agree are far more reliable than one.
Visylix ANPR works exactly this way. It is built in house, reads Indian plates as well as international ones, and needs no country by country setup. A second look, on by default, zooms into each vehicle to catch plates too small to read in the full picture. Because read quality depends on angle, light, speed and plate condition, we prove it on your own lanes during the proof of concept.
Placement matters more than any software setting. The ideal is a dedicated ANPR camera covering one lane, nearly in line with traffic, with its own infrared light for night reads and a fast fixed shutter so moving plates stay sharp. The plate should fill well over 100 pixels of the image width.
Your existing cameras can often do the job where traffic is slow, such as gates and car park barriers, if they record at 1080p or better and look along the lane at a shallow angle. For faster roads, fix the shutter at 1/1,000 of a second or faster to stop blur, or add a purpose placed ANPR camera.
Parking is the biggest use. ANPR logs every entry and exit, recognises staff and subscribers automatically, and tracks how full each lot is. Manual ticketing and repetitive gate checks disappear.
Police and city teams use it to spot stolen and wanted vehicles the moment they pass a camera. Corporate campuses, gated communities, ports and logistics yards use it for vehicle access, visitor and contractor logs, and timing how long vehicles take between gate and dock.
Standalone ANPR forces operators to jump between a plate application and the video system. Visylix runs ANPR inside the VMS. Every read appears as a live event with the camera, a snapshot, vehicle type and colour.
Investigations become a conversation. Ask Radha, the on premise AI copilot, "where did MH12AB1234 go in the last 24 hours?" and it shows every sighting. Because plates are exact identifiers, Visylix links them into plate journeys across cameras, with the travel time between each one.
Plates sit in allow, block, watch, VIP and custom lists. Block and watch matches raise priority alerts. Matching allows for common look alike characters, such as O and 0 or B and 8, so one misread letter does not let a flagged vehicle slip through. Live watches can use patterns, such as every plate starting MH12.
Plate events drive Sudarshan Rules: signed webhooks with retries, MQTT, SIEM and PSIM forwarding, and alarm incidents in the operator console. Your barrier and access control systems keep the final physical decision, so a misread can never open a gate on its own. Every read stays on your own servers and is deleted automatically after the retention period you set.
Accuracy is driven mainly by camera angle, resolution, how many pixels the plate fills, lighting and shutter speed, so a figure from one site rarely transfers to another. Visylix combines readings over several frames and tolerates look alike characters when matching lists, and we measure read quality on your own lanes during the proof of concept.
Often yes, for slow traffic at gates and barriers, if the camera records at 1080p or better, looks along the lane at a shallow angle and uses a fast fixed shutter. For faster roads and reliable night reads, a dedicated ANPR camera with infrared light is better. We review your existing footage during the proof of concept and tell you which lanes need one.
Parking is the largest: automatic entry and exit logs, allow lists and occupancy tracking. Others include stolen and wanted vehicle alerts, campus and gated community access, contractor and visitor logs, and travel time measurement across ports and logistics yards.
Plate reads appear as live events inside Visylix with camera, snapshot, vehicle type and colour. Ask Radha where a vehicle went and see its exact plate journey across cameras with travel times. Allow, block and watch lists apply across every camera, and plate events trigger rules, signed webhooks and MQTT messages to your gate and security systems, all on your own infrastructure.