Automated surface-anomaly detection using machine learning has become an interesting and promising area of research, with a very high and direct impact on the application domain of visual inspection. Deep-learning methods have become the most suitable approaches for this task. They allow the inspection system to learn to detect the surface anomaly by simply showing it a number of exemplar images.

Our research explores a segmentation-based deep-learning architectures that are designed for the detection and segmentation of physical defect detection on reflective surfaces. The case study below shows detection of hail dents on car bodies. The network processes each image and detect the dents in realtime, while surpassing human accuracy, particularly in cases of huge numbers of dents.

SAL v1:

SAL v1.2:

A patent has been filed on the network for reflective surface detection.