Overview

Quality control in lantern manufacturing is crucial for ensuring product safety and efficacy. Traditional inspection methods often rely on manual checks, which can be time-consuming, inconsistent, and prone to human error. The lantern industry needs more efficient and accurate quality control processes to maintain high standards while increasing production efficiency.

Goals

The goal was to develop an automated system for visual quality inspection of lanterns, focusing specifically on the critical aspect of wick positioning.

Solution

We trained a deep learning model to precisely detect and analyze the position of lantern wicks, ensuring they are correctly centered. The Protostar Vision Box, equipped with high-resolution industrial cameras, industrial sensors and an integrated processor is deployed directly on the production line and provides real-time, accurate inspection of wick positioning, significantly improving quality control efficiency and consistency in manufacturing.