Our client wanted to maintain an effective inventory management system through scanning and tracking every item of their inventory. With thousands of items arriving at a time, this nature of monitoring was difficult to set up and maintain with traditional practices, especially if you want to precisely identify product damages, inspect the quality and store the data efficiently. ThinkPalm developed an efficient object detection framework and GUI application for windows by leveraging deep learning capabilities. This solution utilized complex vision tasks such as image tagging, quality inspection, segmentation, or object detection and could be customized to identify specific defects or abnormalities in products. The framework applies object detection at a pixel-by-pixel level and is capable of being implemented into any of the client systems.

Business Transformation

Surface Defect Detection System

Identify/localization of defects with their corresponding label (welding defect, scratch, crack etc.) which are available with databases.
Quality inspection in manufacturing any products viz. automobile, textile, electronics, etc.
A Deep learned trained model can be customized according to the defect to be identified. Purely deep learning approach with digital image processing techniques.