Methodologies for Evaluating Disparity Estimation Algorithms
Iván Cabezas, María Trujillo · IGI Global eBooks · 2012
The use of disparity estimation algorithms is required in the 3D recovery process from stereo images. These algorithms tackle the correspondence problem by computing a disparity map. The accuracy assessment of a disparity estimation process has multiple applications such as comparing among different algorithms’ performance, tuning algorithm’s parameters within a particular context, and determining the impact of components, among others. Disparity estimation algorithms can be assessed by following an evaluation methodology. This chapter is dedicated to present and discuss methodologies for evaluating disparity estimation algorithms. The discussion begins with a review of the state-of-the-art. The constitutive components of a methodology are analysed. Finally, advantages and drawbacks of existing methodologies are presented.