Defect Detection in Turbo Mode

Abstract

Similar to the SDD technique, the ABD technique can also be called adaptive edge-based object analysis. Thus, these two techniques represent different but complementary algorithms which form a new effective class of algorithms for object detection. Each one of these techniques is best suited for the corresponding tasks and shows specific advantages and disadvantages. The SDD algorithm is very effective, reliable, and robust, but it may take a long time if processing of a large image or several images in the inspection of a large area is required. The ABD algorithm is also effective and very fast, but prone to misdetections on high-contrast images. This can happen already when a wrongly detected starting point entails a contour tracing and thus the detection of an object. A combination of all described techniques, however, promises an effective Turbo technique for object detection, working at a high speed on inhomogeneous, high-contrast, and thus very complex surfaces. Such a Turbo technique must be compiled step by step, from a faster to slower step and from lower to higher complexity. In this way, the technique can be accelerated without losing its effectiveness. Beyond that, all steps of the technique can contribute with their adaptivity and flexibility to the Turbo technique. Now we can analyse how and at which point the object detection technique can be accelerated.