SingTown Technology enables robots to "see, walk, and think," maximising competition performance!

In various robot competitions, recognising track boundary lines is a core capability for robots to achieve automatic tracking, path planning, and steering decisions.
Traditional infrared or single-point photoelectric sensors can only detect colour differences and are unable to recognise line angles and curvature information.
To this end, educational robot manufacturers adopt SingTown's OpenMV intelligent camera recognition technology, integrating AI machine vision and image geometric calculation algorithms to achieve real-time recognition, angle measurement, and position tracking of track boundary lines (such as the yellow-grey junction line).
This system can be widely applied in scenarios such as smart car competitions, autonomous driving model cars, and educational programming platforms, providing robots with high-precision visual navigation capabilities.

Educational robot manufacturers utilise SingTown Technology's OpenMV intelligent camera recognition technology to achieve automatic track identification and real-time steering control.

Install a camera at the front end of the robot to overlook the ground track area. The system is equipped with computer vision and image recognition technology, capable of automatically extracting track boundary lines (yellow-grey or black-white dividing lines) through colour threshold segmentation, edge detection, and line fitting algorithms (Hough Transform), while calculating the centre position and angle of the line in real time.

The recognition algorithm outputs include: the coordinates (x, y) of the line segment's centre point, used to determine the robot's offset relative to the track; and the tilt angle θ, used to calculate the robot's steering direction and correction angle.

When the robot strays from the track centre or encounters a sudden change in boundary angle, the system promptly transmits position and angle data to the main control STM32 via serial port or I²C interface, facilitating steering correction and speed control.

The OpenMV embedded vision module maintains stable operation under strong light, shadows, and ground reflection conditions. Leveraging its flexible and efficient on-device AI and edge deployment capabilities, its recognition rate can reach 60fps, enabling true real-time navigation control.

Through this solution, educational robot manufacturers have implemented a machine vision-based track recognition and angle measurement system, enabling robots to "see the track, understand direction, and drive intelligently," significantly enhancing competition performance and intelligent control capabilities.