Video camera-based intelligent transportation systems (ITS) are a broad category from automobile sensing systems, traffic management systems, toll collection systems, and more. ITS can be further broken down into those using automatic license plate recognition (ALPR) or automatic number plated recognition (ANPR).
- The license plate recognition algorithms and software
- The image acquisition technology (camera, optics, lighting, etc.)
These are very diverse skill sets. In this series of blogs on traffic topics, we are attempting to share our experience with the image acquisition challenges.
For instance, the requirements can be drastically different depending on the circumstances. Two extreme scenarios are open road tolling on the highway and parking lot surveillance/automatic payment. While each rely on ALPR, the system specifications to get the necessary image is not the same.
For a parking lot application (slow speed or still image capture), there is no risk of motion blur, and limited obstructions. This translates to an image system with longer integration times and minimized depth of field.
Highway applications are much trickier, there can be motion blur, the potential location of the license plate in the scene is unpredictable, other cars can cause obstruction, limited lighting can be used so as to not affect the driver’s vision, and maintenance is difficult and costly.
This translates to the following camera/image system requirements:
- Short integration times
- High frame rates
- Good sensor selection for necessary sensitivity
- Image processing on the camera to prevent motion blur and adjust for environmental conditions
- Large depth of field
- Reliable trigger
- Robust design and manufacturing
- Complete alignment of the entire electro-optical path
We will continue our discussions on these topics in future posts over the next couple of months.
Image Capture Concerns with Different ALPR Scenarios: