Visibility-enhanced reconstruction of three-dimensional objects under a heavily scattering medium through combined...

In this paper, we propose an effective approach for reconstructing visibility-enhanced three-dimensional (3D) objects under the heavily scattering medium of dense fog in the conventional integral imaging system through the combined use of the intermediate view reconstruction (IVR), multipixel extraction (MPE), and histogram equalization (HE) methods. In the proposed system, the limited number of elemental images (EIs) picked up from the 3D objects under the dense fog is increased by as many as required by using the IVR technique. The increased number of EIs is transformed into the subimages (SIs) in which the resolution of the transformed SIs has been also improved as much as possible with the MPE method. Subsequently, by using the HE algorithm, the histogram of the resolution-enhanced SIs is uniformly redistributed over the entire range of discrete pixel levels of the image in a way that the subimage contrast can be much enhanced. Then, these equalized SIs are converted back into the newly modified EIs, and consequently a visibility-enhanced 3D object image can be reconstructed. Successful experimental results with the test object confirmed the feasibility of the proposed method.