Principles and Applications of RIDED-2D —A Robust Edge Detection Method in Range Images

In computer vision field, edge detection is often regarded as a basic step in range image processing by virtue of its crucial effect. Due to huge computational costs, majorities of existing edge detection methods cannot satisfy the requirement of efficiency in several industrial or biometric applications. This Chapter proposes a novel Rule-based Instantaneous Denoising and Edge Detection method (RIDED-2D) for preprocessing range images. First of all, a new classification is proposed to categorize silhouettes of 2D scan line into eight types by defining a few new coefficients. Subsequently, several discriminant criteria on large noise denoising and edge detection are stipulated based on qualitative feature analysis on each type. By selecting some scan points as feature point candidates, a practical parameter learning method is provided to train and refine the threshold set. RIDED-2D is implemented with three mode algorithms, fastest of which is an integrated algorithm by merging calculation steps to the most extent. Since all the coefficients are established based on distances among the points or their ratio, RIDED-2D is inherently invariant to translation and rotation transformations. For refining the edge lines, a forbidden region approach is proposed to eliminate interference of the mixed pixels. Furthermore, key performances of RIDED-2D are evaluated detailedly, including computational complexity, time expenditure, accuracy and stability. The results indicate that RIDED-2D can detect edge points accurately from several real range images, in which large noises and systematic noises are involved, and the total processing time is less than 0.1 millisecond on an ordinary PC platform using the integrated algorithm. Comparing with other state-of-the-art edge detection methods qualitatively, RIDED-2D exhibits a prominent advantage on computational efficiency. Thus, the proposed method is qualified for real-time processing in stringent applications. Besides, another contribution of this chapter is introducing CPU clock counting technique to evaluate the performance of the proposed algorithm, consequently, the technique suggests a convenient and objective way to estimate algorithm’s time expenditure in other platforms.