Depth maps generation through thresholding in a robotics system
The project described in this paper aims to develop a simple but efficient method for mapping the depth of an object used in an industrial application. In order to produce the depth and cross-sectional images of an object the intensity threshold for converting the grey image to a binary image is varied. This operation will not produce a satisfactory result when the object is observed in air. The result will be however significantly different when the object is located in a colour liquid. Since the refraction of the light in the liquid varies with the depth, the intensity of the reflected light alters with the height of the object. Results of the method applied to a complex casting are presented showing the method's effectiveness.
Please refer to publisher version or contact your library.