Image Recognition: Autos and more about OpenALPR

INSIGHTS

  • OpenALPR is a cloud-based vehicle recognition service that supports vehicles made in several countries, including Thailand, India, and Indonesia.

OVERVIEW

After thoroughly going through our sources and databases, we were unable to find information regarding car-recognition technology in Asian-Pacific countries similar to how Blippar and Orpix recognize vehicles from the U.S. market. As of this time, market research agencies have not captured this information in their reports. The reports they do release focus mainly on general aspects, such as a global focus of Vehicle License Plate Recognition systems (VLPR). While information regarding API technologies surely exists somewhere, it remains scarce in the public domain. Among the sources we found, almost all of them focused solely
on VLPR rather than vehicle recognition. API technology is fairly new, even in the West, so our lack of results may even be the cause of a market too early in its development to yield the desired information. Below, we have detailed the methodology used to search for API information as well as some other findings that may be of use going forward.
METHODOLOGY
We first searched for information regarding companies that offer auto/car recognition API using AI or computer vision technology in Asia-Pacific countries. We utilized resources such as Crunchbase, Owler, and Bloomberg, and we were able to get a lead on one company named OpenALPR. Through the company's website, we were able to find details on the company and the Asian countries it services (see "Helpful Findings").

Once we felt we'd thoroughly investigated OpenALPR, we looked for information highlighting some cars the technology recognizes from the Indian, Indonesian, Thai, and Asian-Pacific auto markets, but the search turned was fruitless. We tried a similar approach through Techcrunch, Ovum, Frost, Sullivan, and other technology magazines and research websites but failed to find any new information.

In our final attempt, we switched back to gathering information on API tech in Asian-Pacific countries by looking through sources such as Red Newswire and Transparency Market Research, but any resources we found had a global lens. Furthermore, most of these reports covered license plate recognition rather than vehicle recognition.

HELPFUL FINDINGS
The best piece of information we gathered was regarding the company OpenALPR Cloud API. OpenALPR is a cloud-based web service that identifies vehicles in images and returns vehicle color, make, model, and body type as well as license plate data. They service several Asian-Pacific countries including Japan, China, Australia, United Arab Emirates, Indonesia, Saudi Arabia, India, South Korea, and Thailand. The tool can be used at several levels with varying prices and capabilities.

Any other sources we found talked strictly about vehicle plate recognition, but we have included them here in case something useful might come of them. Red Newswire published a report on Vehicle License Plate Recognition systems that may be useful. It is locked behind a paywall, but it claims to have information on Southeast Asia, Japan, India, and China from 2012 to 2016 with a forecast to 2022. There is another report about automatic number plate recognition published by Transparency Market Research, but this, too, is behind a paywall. The report discusses strategies in the market and its high degree of fragmentation at present.

Image result for Image Recognition: AutosWe also found two pieces of software of note. The first is Asia Vision Technology, Ltd (AVT). It calls itself "the world's leading technology and solution provider of license plate number and container number recognition." Its involvement in the development of technologies such as Optical Character Recognition suggests that it may be privately involved in car recognition API or may involve itself in the future. The second software we found is Sighthound, which has recently "announced the release of new vehicle detection and recognition engine for its Sighthound Cloud API service." It has both vehicle and facial recognition, but we did not include it with OpenALPR because all of its Asian-Pacific vehicles examples are brands recognized in the West.

CONCLUSION
Most likely due the novelty of the technology, we were unable to identify any artificial intelligence or computer vision technologies with auto/car recognition capabilities. The one that we were able to identify is OpenALPR, which can be utilized with Western cars as well as many Asian-Pacific vehicles, including brands from Thailand, China, and United Arab Emirates. Even reports concerning vehicle license plate recognition technology are hard to come by or locked behind paywalls, so it is likely that vehicle recognition tech, which is understandably more difficult than the former technology and probably further behind, simply doesn't have the research available to give a comprehensive answer.

SOURCES

Post a Comment