Virtual Grid Mapping for Visual Size Measurements

This article proposes a novel, more robust method for single-image visual size measurements, named virtual grid mapping (VGM). VGM requires just a single image as input; however, it does not require any prior information concerning the scene, like the horizon line or reference objects. Unlike current methods, the VGM approach uses a grid of virtual 3-D points projected to the 2-D image plane, along with the fusion of probable depth values indicated by 2-D–3-D point correspondences; it manages to reduce the uncertainty originating from the calibration and the positioning of the camera; thus, providing more accurate measurements. Given the geometric properties of the camera, a novel approach of VGM is that it automatically generates and projects a grid of virtual 3-D points to the 2-D image plane, enabling the establishment of approximative correspondences between 3-D points of the real world and 2-D points of the image plane. Then, by considering these initial 3-D–2-D point correspondences as known, a range of possible depth values is estimated, and through an adaptive fusion process, the size of an object of interest can be accurately approximated. Another advantage of VGM over the state-of-the-art deep learning (DL)-based methods is that it requires a simple training process. Experiments performed on simulated and real image datasets, captured in controlled conditions and in the wild, show that the mean absolute percentage error (MAPE) of the proposed method ranges between 3.13% and 11.66%.