In the scope of three-dimensional (3D) measuring methods, structured light profilometry (SLP) and shape from focus (SFF) methods are based on distinct optical setups with their own advantages, such as good reliability for SLP and extended depth of field for SFF. This article proposes to adapt an SFF method on a device configured for SLP which constitute a first step towards the fusion of SFF and SFP methods. The proposed SFF method remains valid although the optical axes of the projector and camera are not aligned with the SFF translation direction. This configuration is unconventional for SFF; therefore, each point of the scene is no more static on the captured images during the translation process. To overcomes this phenomenon, each local area of the scene along the set of captured images is tracked using calibration data. A large part of the calibration steps applied to the proposed SFF method are similar to the ones used for SLP, which are based on homography calculation. Furthermore, during the measurement process, the patterns projected for the proposed SFF method to measure the focusing are identical to the ones projected for SLP. To demonstrate the validity of the method, experimental results are provided with depth profile comparison between SLP and proposed approach of SFF.
|