This automated system detects various mounted printed circuit board components by analyzing images of printed circuit boards. Printed circuit boards are critical components of nearly all modern electronic devices, and the printed circuit board market is projected to reach $75.52 billion by 2026. Printed circuit boards are mass-produced following the validation of a printed circuit board prototype, but prototype components must be thoroughly validated and a bill of materials needs to be produced prior to mass-production. The most commonly used method for prototype validation requires human subject matter experts to manually validate components and produce a bill of materials, making this process time-consuming, labor-intensive, and expensive.
Researchers at the University of Florida have developed a computer-implemented system that automatically detects printed circuit board components, an essential step for generating a bill of materials in an optical assurance pipeline. This allows for faster and more accurate detection of components and bill of materials generation.
Automatically detects printed circuit board components useful for reverse engineering, hardware assurance, or industrial assessment
This automated system for detecting printed circuit board component defects and generating a bill of materials, uses an algorithm to assess images of the printed circuit board and then estimates component locations. This method uses an angled directed light source to cast shadows on the printed circuit board. The shadows are stored as binary images and a proximity threshold is used to ensure only one component is assigned one shadow. The components are extracted from the binary image and compared to an unlit reference picture to validate the location of the component. This process can be used to quickly and accurately generate a bill of materials and can detect printed circuit board component defects.