News | FF-GROUP

I want to share some thoughts about EDGE solutions from a practical and commercial point of view. The following situations are best for using on-camera software:

1 The commercial factor. Network objects with dozens of cameras located at points at significant distances (for example, gas stations throughout the country, region). Installed on cameras, EDGE (LPR / MMR app) saves project money on servers and data transmission infrastructure. It is enough 4G network and 1 server per country. Wherein the network structure is well-built and cheaper.

2. Scalability. When creating large intelligent systems, the use of on-camera applications ensures their rapid scalability by simply adding a new data source (data sensor). The cost of the system practically does not change, and the amount of received traffic information increases several times.

3. Quick start of the system. A quick start of data generation (just a couple of minutes) from the moment the IP camera with built-in EDGE application was installed. There is no need to add servers, just install the camera and connect to the system.

4. Simplicity and mass character. On camera LPR / MMR analytics has always been very complicated for producing information and an exclusive feature in ITS projects. Today FF Group was able to transmit complex algorithms for recognizing all vehicle parameters on a regular IP camera into a simple and convenient software. Now, sophisticated EDGE video analytics is moving into the niche of simplicity and mass commercial use.

5. A stable system. Sustainability lies in the low risk of failure of their data sensor. It is enough to replace the camera with the EDGE app. In this way, the collection data point is quickly restored at a low cost.

Short summary: With EDGE app ITS projects are architecturally simple, easily scalable, the data generation time is short, and at an affordable price. In a nutshell, it’s convenient!

1. LPR applications on EDGE have already taken a strong position in the CCTV market with a sufficient number of players. It has become not just nice technology, but a commodity.

2. Strengthen the LPR App with vehicle make, model, color and type recognition (MMR). It should be on EDGE with high quality. All FF Group products will be strengthened with MMR next year.

3. It is important to bring EDGE to a global level. The US market starts looking for EDGE solutions, and the Middle East is waiting for high-speed and high-quality app for large traffic projects.

4. To scale EDGE globally, all world number plates have to be described. The best way to do this is the synthetic databases for testing algorithms, which allows you to connect new regions in a short period.

5. AI is transforming. Today it is not only training on large volumes of static data (machine learning), but also the active use of theoretical descriptions of various AI models.

6. The upcoming trend is producing useful extra information on EDGE based on measured data (for example, counting city transit time). This will simplify the operation of traffic and video management systems.

Let’s move on EDGE in a new way in 2021!

Can real license plate recognition be more than 95%, for example, 99%? And what does the 99% recognition accuracy rate mean for the customers? Recognition accuracy is an excellent marketing indicator of the LPR software quality, understandable to the customer.

It can be argued that if the integrator follows all the instructions for the camera set up and its installation, then the software recognition accuracy will be at least 95% in all countries. In some cases, with proper camera installation and operation, it can reach 99%. Which indicator should you believe?

Let’s explain it with Russell’s teapot story. And also look through the other important indicators the LPR software quality.

More details are in our blog.

What does 99% recognition accuracy mean to you?