Evaluation Of Subpixel Line And Edge Detection Precision And Accuracy

An extensive analysis of the quality of the extraction results of a subpixel line detector and a subpixel edge detector is carried out. The localization quality of line and edge points in the presence of noise is analyzed analytically, and new formulas describing this relationship are derived. Tests on synthetic noisy images show the formulas to hold very well in practice. Experiments on real images show that subpixel accuracy better than one tenth of a pixel is possible in typical industrial inspection or close-range photogrammetry tasks. 1 INTRODUCTION The analysis of the performance characteristics of a feature extraction algorithm is very important (F orstner, 1996). First, it makes an algorithm comparable to other algorithms, thus helping users in selecting the appropriate method for the task they have to solve. Second, it helps to identify breakdown points of the algorithm, i.e., areas where the algorithm cannot be used because some of the assumptions it makes are violated. Ther..