Feature Extraction and Seam Recognition Techniques
The previous chapter discussed the various techniques available for sensing and measurement of the surface geometry. This chapter provides details on processing the vision data for extracting and recognizing the seam’s features in the image and subsequently modeling the seam environment. In a broad sense, the images generated as a result of the signal transduction by vision-based systems can be classified as intensity images or range images. Intensity images are a rectangular array of intensity values corresponding to spatial points on the surface being measured. On the other hand, range images generated using position sensitive photo-detectors can be described by an array of range values (distance) corresponding to spatial points in the scene being measured. These range image arrays are rectangular when the laser beam can be moved along two axes and linear in the case of synchronized scanning systems where the beam is scanned along a line.
KeywordsFeature Point Weld Seam Range Image Certainty Factor Seam Tracking
Unable to display preview. Download preview PDF.