Machine vision solutions

With the rapid development of machine vision, the application of artificial intelligence in various fields is becoming more and more important. It is gradually used in security monitoring, medical imaging, robotic industrial vision, automated driving, drones and other industries. With the development of image recognition technology and deep learning With the continuous development of algorithms, the object recognition and ranging of machine vision are constantly deepening and expanding.

The machine vision object recognition and distance detection solution uses high-performance open source motherboards and high-definition cameras. It combines machine vision and deep neural network technology. It not only identifies the type of objects in front of the camera, but also calculates the distance between the obstacles ahead and the camera. The camera distance detection accuracy error is 0.05m. Compared with the traditional machine vision recognition, it can be better applied to various intelligent fields.

Recommended solutions:

     LQ3399Pro open source motherboard + monocular camera/binocular camera

Feature description:

      1. High-performance AI processor

Using RK3399Pro high-performance AI motherboard, ARM six-core processor architecture, clocked at 1.8GHz, quad-core graphics processor Mali-T860 MP4, integrated neural network processor NPU, computing power up to 3.0Tops, compatible with multiple AI frameworks.

2. Configure HD camera

Equipped with a high-listing camera/binocular camera, it can quickly realize the recognition of objects and ensure that the picture is clear. The detection frame rate of the monocular camera reaches 25fps@640*480, and the distance detection accuracy error is 0.3m; the detection frame rate of the binocular camera reaches 15fps@640*480, and the distance detection accuracy error is 0.05m.