A conventional toll plaza gives the system time to identify a vehicle as it slows near a booth. Open road tolling removes that pause. The vehicle may pass the gantry at highway speed, change lanes near the capture point, or travel beside a truck that blocks part of the view. Every later step depends on the image recorded during that brief movement.
Open Road Tolling (ORT) and Multi-Lane Free Flow (MLFF) are often used interchangeably. ORT refers to any toll system without physical barriers collecting tolls, while MLFF describes specifically the ability to collect tolls from vehicles moving in multiple lanes, such as those changing lanes under an overhead structure (gantry).
India’s rollout illustrates how quickly this technology is scaling. The country implemented its first MLFF tolling system at the Choryasi Toll Plaza on the Surat–Bharuch section of NH-48 in Gujarat on May 1, 2026, processing around 41,500 vehicles on launch day. In this system, FASTag and Automatic Number Plate Recognition are integrated, but captured images remain the primary transaction record when tag reads are unsuccessful, contested, or associated with the incorrect vehicle.
In the first part of this three-part blog series, you’ll learn how imaging works in ORT and MLFF, the various challenges faced along the way, and how to overcome them.
Understanding Gantry Geometry and Multi-Lane Coverage
Overhead vs. Side mounting
Overhead mounting reduces occlusion from adjacent vehicles but flattens the plate angle, reducing pixel detail. Side mounting offers a stronger plate angle, but risks occlusion. For instance, a truck in the nearest lane may hide a car beside it.
Field of View and lane coverage
Field of view (FOV) involves a direct trade-off. Wide FOV covers multiple lanes with fewer pixels per plate, while narrow FOV delivers better plate detail but may require multiple cameras per gantry. Planned overlap between neighboring views protects coverage during lane changes, but must be coordinated to avoid duplicate records.
The gantry also hosts RFID readers, edge processors, and communication systems. Camera coverage, RFID read zones, and the capture point must align so that the image and tag event refer to the same vehicle.
Role of Pixel Density and Plate Readability
Pixel density determines how much plate detail the camera captures. Pixels per foot (PPF) or per meter (PPM) describe this: divide the camera’s horizontal resolution by the roadway width visible at the capture point.
For reliable OCR, a one-line plate should occupy roughly 100 to 150 pixels over its width. Confirm the final target with your OCR supplier and local plate formats, as fonts, reflective materials, and two-line layouts can change the requirement.
Wider coverage reduces pixel density; narrower coverage improves it but requires more cameras. Therefore, resolution can’t be assessed in isolation. Lens choice, mounting angle, and lane position all affect the pixels available to OCR. So, always check outer lanes and overlap regions. Please note that a design that works at the lane center may fail at the edges.
Imaging Challenges of Vehicle Diversity
Free-flow tolling must handle diverse vehicle types because each comes with unique plate positions, sizes, and reflective properties.
| Vehicle Type | Challenge |
| Motorcycles | Smaller target, varied angles, and risk of plate blockage |
| Oversized trucks | Plates at different heights; can block adjacent vehicles |
| Trailers | May carry separate registration from the tractor |
| Temporary/foreign plates | Different fonts, colors, spacing, and reflective materials |
| Damaged plates | Bent, dirty, or obscured plates can fully remove readable characters |
Meeting IHMCL Specifications for ANPR Performance
For Indian MLFF deployments, camera performance is a contractual requirement. Under IHMCL specifications, ANPR cameras must meet benchmarks such as:
Hence, the camera is transaction-critical. If it fails these specifications, then the camera can’t be deployed in Indian MLFF projects.
Global vs. Rolling Shutter for High-Speed Capture
Highway speed leaves little time for exposure. If the exposure remains open for too long, the plate moves far enough to soften character edges. If it is shortened without enough illumination, noise can reduce contrast. Sensor sensitivity, lens aperture, illumination, gain, and exposure time must work together so that the plate remains readable while the vehicle passes the capture point.
A global shutter exposes all pixels at the same moment, preserving the geometry of a moving plate. A rolling shutter exposes rows in sequence. Vehicle movement between the first and last row exposures may create slanted plate edges, skewed characters, or a distorted outline.
Controlled lighting and short exposure can reduce this risk, but a global shutter is generally better suited to highway-speed capture.
Synchronization is also important when several cameras or an infrared illuminator are used. The exposure window must coincide with the light pulse, and each camera needs a common timing reference. Poor synchronization can leave one view dark or record different vehicle positions in images expected to describe one transaction.
Finally, trigger position and timing decide where the vehicle appears in the frame. A late trigger may cut off the plate or miss the vehicle after it passes the reading zone. An early trigger may record the plate while it is too small or near the image edge.
How to Ensure Day/Night Imaging Performance
Daylight and high-contrast scenes
A tolling camera may face direct sun, gantry shadows, headlight glare, wet-road reflections, and rapid brightness changes. High Dynamic Range, or HDR, helps retain detail in bright and dark regions within one frame. It still requires scene-based tuning. Excess exposure can erase reflective plate characters, while low exposure can hide dark vehicles and context.
Night capture and varied plate types
Infrared illumination supports short exposure after dark without depending on vehicle headlights. The illumination angle, output, and timing must deliver enough light at the intended capture point without creating a bright reflection that erases the characters. The pulse also has to coincide with the sensor exposure.
Dual-sensor design
A dual-sensor design can use one imaging path for visible color and another for monochrome or near-infrared capture. The color stream supplies vehicle evidence, while the infrared stream can produce stronger plate contrast at night. The two views need aligned coverage and synchronized timestamps so they can be associated with the same vehicle.
Non-standard plates complicate visible and infrared capture. Temporary plates, faded paint, decorative frames, reflective coatings, unusual colors, and two-line formats respond differently to each light source. Testing should include local plate types during day, dusk, night, rain, glare, and wet-road conditions. Tolling needs consistent image quality through the full operating day.
The Need for Environmental Reliability of Cameras
Environmental protection is part of the optical design because water, dust, movement, or a changed camera angle can reduce plate readability even when the sensor and lens were selected correctly.
- Ingress protection should cover the enclosure, optical window, connectors, cable entry points, and seals used in the installed assembly.
- Vibration resistance should keep the camera angle, lens position, and focus from changing under traffic-induced movement.
- Software-defined lane reconfiguration should support updates to regions of interest, virtual trigger lines, and capture zones when lane usage changes.
It is important to understand that environmental ratings must not replace routine inspection.
- Dust, mud spray, insects, and residue on the optical window lower contrast and scatter illumination.
- Loose brackets alter the camera angle, while a blocked window can remove the plate from view.
Regular washing, focus checks, mounting checks, and sample-image review should be included in corridor operations. In the third blog of this series, you’ll get more in-depth information about this topic.
Why Every Captured Frame Has a Revenue Consequence
In free-flow tolling, the camera gets one brief chance to record a vehicle as it enters the capture zone. The image must then remain usable while the vehicle is moving, which means shutter control, exposure, synchronization, and trigger timing have to work as a single capture sequence.
The revenue consequence of the effectiveness of this imaging performance can be significant. Using an illustrative average toll of $1.50, a 1 percent miss rate on a corridor carrying 50,000 vehicles each day would create 500 missed transactions daily. In this illustrative scenario, the annual revenue leakage would reach **$273,750**.
Pixel density is a budget issue, as much as an imaging calculation. A readable plate image completes the optical stage, but it does not complete the transaction.
In the second blog of this series, you’ll get expert insights into how the system validates the image against the tag, vehicle class, and payment event before the transaction can be accepted.
e-con Systems Offers AI Vision Solutions for MLFF and ORT Setups
Since 2003, e-con Systems has been designing, developing, and manufacturing OEM and ODM cameras. Drawing on two decades of OEM camera design experience, we have developed an edge AI vision solution purpose-built for MLFF and ORT gantry environments. Its advantages include:
- Simplified tolling with reduced infra needs for fast rollouts
- Low-power envelope for solar sites
- >99.95% detection and classification accuracy (under controlled test conditions across varied lighting, weather, and vehicle types)
- Support for vehicle occupancy detection, speed and violation analysis, dynamic pricing, and reservation-based lane allocation
Recently, we worked closely with a North American MLFF operator to offer an edge AI imaging system for Multi-Lane Free-Flow tolling.
Please visit our Camera Selector Page to browse our full portfolio.
Looking for the ideal imaging solution for any smart traffic system? Please get in touch at camerasolutions@e-consystems.com. Our experts are readily available to help you find the right solution.
FAQs
Why does camera design matter in ORT and MLFF systems?
The camera creates the primary visual record of each vehicle. Poor placement, low pixel density, weak timing, or unsuitable lighting can reduce plate readability and affect toll collection.
What is the difference between ORT and MLFF?
ORT refers to barrier-free toll collection. MLFF refers to toll collection from vehicles moving through multiple lanes, including vehicles changing lanes near the gantry.
How much plate detail is needed for reliable OCR?
Many license plate recognition systems need roughly 100 to 150 pixels over the width of a one-line plate. The final target depends on plate format, font, material, and OCR requirements.
Why is a global shutter preferred for high-speed tolling?
A global shutter captures all pixels at the same moment. This reduces skew and character distortion when vehicles pass the camera at highway speed.
How can tolling cameras maintain image quality during day and night?
HDR helps manage bright and dark areas, while infrared illumination supports short exposure after dark. Dual-sensor imaging can also provide color evidence and stronger plate contrast.
What are the IHMCL ANPR camera specifications for Indian MLFF projects?
The specifications include 2 MP per lane, global shutter, >50 dB SNR, H.264 encoding, IR night vision, Bad Pixel Correction, Edge Enhancement, and mandatory confidence-level reporting (>95% for automated transaction processing).
Dilip Kumar is a computer vision solutions architect having more than 8 years of experience in camera solutions development & edge computing. He has spearheaded research & development of computer vision & AI products for the currently nascent edge AI industry. He has been at the forefront of building multiple vision based products using embedded SoCs for industrial use cases such as Autonomous Mobile Robots, AI based video analytics systems, Drone based inspection & surveillance systems.