DCF: Disparity computing framework for stereo vision systems
Gabriel da Silva Vieira, Júnio César de Lima, Naiane Maria de Sousa, Fabrizzio Alphonsus Alves de Melo Nunes Soares · Software Impacts · 2022
Disparity maps are vital components of stereo vision systems as they encode the displacement of two or more images. However, previous works provide only a few implementation details, suggest processing steps that are not very well defined, and the software design is rarely discussed. Conversely, DCF applies the main components of a stereo vision system and integrates them to promote disparity map construction. As a result, DCF algorithms can be parameterized or executed with previously defined configurations. Thus, DCF outputs can be directed to different applications, such as benchmarking proposals, computer and robotic applications, triangulation, and 3D reconstruction.