Amped Software is proud to announce the launch of DeepPlate, their new online platform designed for AI-powered reading of severely degraded license plates (for investigative purposes only).
License Plates: an Evergreen Challenge
It’s no secret that reading a poor-quality license plate is one of the most common challenges for forensic video analysts and investigators. Despite the constant increase in the spatial and time nominal resolution of surveillance cameras, it is still very common to deal with heavily compressed or poor-resolution license plates that are hard to read.
Amped FIVE’s Perspective Super Resolution feature is often fundamental in enhancing low-resolution license plates. Additionally, the Frame Averaging feature can significantly improve the quality of a decently sized but very noisy license plate. However, it should be noted that Amped FIVE does not possess the capability to interpret the characters on a license plate directly. Introducing DeepPlate will allow users to evaluate and compare their results with an “impartial reading” of the plate.
What Is DeepPlate
Amped Software has developed DeepPlate, a deep-learning-based algorithm designed to decypher license plates that are commonly affected by issues associated with surveillance systems. These issues include perspective distortion, poor resolution, optical and motion blurring, and compression noise.
Amped Software achieved this by training a dedicated neural network with millions of synthetically generated and distorted license plates for different countries/states. The process of generating synthetic license plates becomes fairly easy once the specific font, spacing, and allowed character structure are identified (customized license plates are not considered at this stage). This approach allowed the company to avoid scraping license plates from the web, thereby eliminating any concerns regarding data provenance and privacy in their model.
Amped Software has documented the technical aspects of their system in a scientific paper, which includes experimental validation on Italian license plates. Although the model has undergone improvements and additional countries have been incorporated since publication, the fundamental concept remains consistent.
Currently Supported Countries
The currently supported countries are: France, Germany, Italy, Netherlands, Spain, Sweden, United Kingdom, and United States of America. For countries with multiple license plate formats, the supported formats are displayed during the license plate upload phase.
DeepPlate Access Conditions and Data Storage
DeepPlate is currently offered at no extra cost to Amped FIVE users who have an active subscription license or an SMS plan. To qualify, users must be part of an institution located in one of the supported countries listed above. The service is accessible from the Amped Support Portal.
Each institution meeting the requirements above will be gifted 50 DeepPlate uses a month for every Amped FIVE license (and for each seat, in case of multi-seat licenses). Please be aware that this usage cap is shared among all users associated with that license. The company doesn’t impose any constraints on how usage is split among users.
Data Storage
Amped Software will only hold your data on its servers for the short time needed to process them. It will not retain images nor use them to train its AI model. The company simply processes the images through DeepPlate and then deletes them, ensuring a strong commitment to user privacy and data security.
Final Remarks
Amped Software’s research team has dedicated several years to the development of DeepPlate and is pleased to share the results of its efforts.
If there are suggestions for improvements, the company welcomes feedback via its contact page. For detailed instructions on how to access and use DeepPlate, read the dedicated blog post.
