AI SPECTR – solution for non-metallic material surface inspections
A hardware and software platform to support operators in inspecting surface defects in non-metallic materials.
Defect detection is performed by analyzing the video stream using artificial intelligence and machine learning (deep learning), as well as pre-trained models.
The developed hardware-software complex (PAC) will provide first aid to the operator, using a set of models to identify the “rainbow” defect and works according to the following algorithm:
- a special light source placed in the hardware illuminates the area under investigation
- the camera of the device transmits the video stream of the illuminated area to the server, where it is processed using machine learning algorithms for possible defects.
- The processed video stream is transmitted back to the PAC screen.
The entire process is fully automated using:
- custom mobile inspection software
- computer vision and machine learning algorithms
- cloud server for data collection and processing.
