Video analysis solution.We are developing data and video analysis solutions necessary for smart cities and autonomous driving. / Smart transportation (accident prediction field)
1. Prediction of collision risk.
It is a solution that predicts and warns of collision risks by predicting future situations.
The vehicle and pedestrian are analyzed through a product called a butterfly box, and a small circle as on the right has moved, A large circle represents the future prediction path. There is a risk of an accident when a large circle overlaps.
Accordingly, the risk level is divided to warn the vehicle and pedestrians.
The vehicle is warned by an electronic display, and the pedestrian is guided in the form of a voice and a floor logo. You may receive a warning. Using the above image analysis technology, even for the visually and physically handicapped, I'm helping them. For the visually impaired or physically handicapped, if an unexpected situation occurs or objects that did not exist, It's hard to deal with. However, using Blue Signal's situation prediction technology, real-time risk information It delivers and helps you cope with dangerous situations in advance. Next time, I'll use a butterfly box on my smartphone. Interworking with an autonomous vehicle, the butterfly box delivers the information predicted to other devices.
Other devices are trying to cope with future risk situations through the information delivered by Butterflybox.
Butterflybox is currently conducting a demonstration project in Sejong City.
2. Butterfly box (artificial intelligence-based accident prediction CCTV)
Butterfly Box: Artificial Intelligence CCTV Predicting Accidents
Butterflybox refers to a device that predicts an accident by predicting and warning various dangerous situations occurring at an intersection while monitoring the situation at its own intersection in real time.
It is possible to prevent possible accidents at intersections by predicting dangerous situations that may occur after one to two seconds.
Predicting dangerous situations.
- Jaywalking pedestrian.
- Traffic violation vehicle.
- All moving objects, including pedestrians, when turning right) Crash situations due to undiscovered
- Warning of collision situation between vehicles due to lane invasion.
1)Butterfly boxes identify objects in pedestrians, vehicles, motorcycles, kickboards, electric wheelchairs, bicycles, etc. passing through intersections and identify the intention of these objects to do.
2) It mainly judges the direction of the final target point or the shape of the joint on the object, or warns moving objects by predicting the collision point.
3. Innovation and differentiation of management functions in butterfly boxes.
Maintenance is convenient and there will be no future costs.
- Automatic registration and management program linkage automation when installing a new butterfly box.
- One-click upgrade when upgrading a new version (maintenance costs nothing).
- Convenient user interface.
1) Warning method.
Warning (pedestrian) through logo projects and hyper-oriented speakers (focusing only on relevant pedestrian locations).
An electronic display board in front of the vehicle
when turning right directly to the vehicle.
An electronic display board
in front of the vehicle when turning right directly to the vehicle.
[The type of warning that pedestrians see]
3) Butterfly box package system configuration.
- Go to the homepage exclusively for Butterfly Box.
4. Applied Reference: Sejong City
- Price: Installation of 1 or 2 butterfly boxes per intersection Basic price: Discussion
- Effect: 15% reduction in traffic accident rate (number of collision accidents between vehicle and pedestrian, jaywalking of pedestrian, traffic accident due to violation of vehicle signal)
1) The actual intersection screen difference interpreted by the butterfly box.
Butterflybox is a solution to reduce traffic accidents and will develop into a valuable solution to protect our growing children, especially .
5. Real-time video analysis solution.
A real-time obstacle recognition solution for the disabled and the weak.
- When collecting pedestrian path information based on artificial intelligence-based real-time obstacle recognition, obstacle information is transmitted.
The purpose is to provide information when creating risks or dedicated maps by identifying obstacles and locations in real-time when walking by the weak or the weak in traffic.
By analyzing the image to be learned by artificial intelligence, it separates and analyzes various objects in the image.
Object recognition type: We are learning various objects that may occur on the sidewalk, and we plan to expand them to recognize various types of obstacle information, including construction information and triangles.
6. AUTO LABELING
The artificial intelligence engine automates the main labeling of videos for learning.
It innovatively reduces the process and time of artificial intelligence learning by automating the labeling process that has been done manually for artificial intelligence learning so far.