Unknown scale objects recognition using massively parallel processing with CUDA

The paper presents a method of irregular patterns creation and matching in an example computer vision system designed for unknown scale objects recognition. The matching process is based on the Hough transform for irregular objects with a parameter space defined by translation, rotation and scaling operations. The high efficiency of the technique allows for poor quality or highly complex images to be analysed. The technique may be used in robot vision systems with unknown scale of the scene, identification systems or for image analysis, directly on grey-level images. In order to guarantee the performance the process of patterns matching is supported by massively parallel processing with the CUDA technology.