Vision Tech

We have developed our own ultra-high-resolution cameras, which are central to our high-accuracy recognition technology. To achieve precise positioning, we handle the entire development process from lens design and lighting design to our proprietary vision processing algorithms. Furthermore, our camera units are designed not only for installation in machines but also for use in external systems. Using the same camera units consistently throughout the entire process from data generation to part placement ensures stable, high-accuracy placements.

To achieve high-speed vision processing, precise control of the camera's imaging timing in synchronization with high-speed axis movements is essential. Furthermore, since this process is performed for all picked parts, the system is designed on the assumption that imaging and vision processing will always operate in a continuous loop. These series of control functions are designed to enable the SMT pick and place machine to perform all operation from part pickup to placement without stopping, thereby achieving both faster processing speeds and the same high placement accuracy as conventional systems.

To achieve high-quality, high-accuracy placements, it is essential for SMT pick and place machines to perform recognition from multiple perspectives. For example, even for a single part being placed, various recognition processes are performed, such as detecting the part top surface, checking the part stance, positioning for placement, checking whether the part remains on the nozzle after placement, and verifying that the part has been placed. In addition, SMT pick and place machines perform vision processing on all target objects within the machine, to ensure stable, high-quality, and high-accuracy placements.