[AI Image Inspection Case] Confirmation of Proper Insertion of Terminals

By teaching the system what to look for, we can achieve higher precision in inspections!

The demand for inspection automation due to labor shortages has been increasing year by year, and amidst this, AI-based inspection systems have also diversified. With the theme of "easy operation for on-site users," we released an all-in-one and simple-to-operate AI image inspection software in 2020, and we have steadily increased our implementation achievements. By circling the items you want to find (OK/NG) with a mouse and clicking the "Learn" button, the inspection machine grows to enable easy pass/fail judgment. For object recognition, we use transfer learning and high-accuracy models, allowing for higher precision inspections by teaching the system to recognize NG items and the objects you want to find. Using the inspection function of 'DeepSky,' we labeled (named) correctly inserted terminals as "Normal" and terminals with insertion errors as "Abnormal," enabling accurate identification. In actual inspections, we set it up so that if even one "Abnormal" is found, it results in a NG. (Settings can be freely adjusted, such as determining what constitutes OK/NG based on the number of items found.) [Software Used] ■DeepSky *For more details, please refer to the PDF materials or feel free to contact us.

basic information

**Features of DeepSky** - An all-in-one AI image inspection software that is easy to operate and simple to use. - Capable of performing "data collection → learning → judgment" on a single PC. - Simply enclose the area you want to find in the captured image and click the "Learn" button. - Easy to set up, demonstrating strength in judgment under complex and variable conditions. - Compatible with ultra-high-resolution cameras at tens of millions of pixels and multi-camera setups. - A free evaluation service is also available. *For more details, please refer to the PDF document or feel free to contact us.*