'Mekikibaito' is an AI product anomaly detection system that turns the inspection line into a photo studio
There are many different types of AI-based product anomaly detection systems. Mekikibyte , developed by Chutzpah Inc., is committed to taking high-quality images, and adjusts the type of light source and the color of the conveyor belt to suit the target product, allowing it to take images suitable for analysis by AI.
Mekikibaito: Visual Inspection and Quality Control AI for the Manufacturing Industry | Hutzper Inc.
I found a booth for Chutzpah Co., Ltd. at the food-related trade fair ' FABEX Kansai 2024. ' A demo model of 'Mekikibaito,' a quality control system that uses AI, was on display.
Anomaly detection systems using AI generally use a mechanism to 'take product images with a camera and analyze the images with AI,' but depending on the factory, it may be difficult to obtain high-quality images suitable for analysis by AI due to factors such as the brightness and position of the lighting. Mekikibaito is committed to 'obtaining high-quality images,' and customizes the colors of the inspection line lighting and conveyor belt to obtain high-quality images without being affected by factory lighting, etc.
The light source of the demo model looks like this. It is designed to illuminate the subject with a constant brightness from the same direction as the camera.
Apparently the color of the conveyor belt can also be customized to suit the requirements. The conveyor belt of the demo machine was white.
Here's what it looks like when the food passes under the camera. You can see that the food is brightly lit.
The captured image is below. It is possible to record very clear images. By turning the inspection line into a photography booth, Mekikibaito can record high-quality images suitable for analysis by AI.
For more information about Mekikibaito, please see the link below.
Mekikibaito: Visual Inspection and Quality Control AI for the Manufacturing Industry | Hutzper Inc.
https://hutzper.com/mekiki-baito/
Related Posts: