AUTOMATIC FAÇADE MOSAICKING USING MONOCULAR IMAGES
Keywords:
Close-range photogrammetry, image processing, facade mapping, image mosaicking, camera calibration, vanishing point geometry, 3D reconstruction, texture mapping, least squares adjustment, computer vision.Abstract
This paper presents an automated approach for building facade
reconstruction and texture generation using monocular close-range image sequences. The proposed method focuses on generating seamless facade textures by processing images captured from ground-level camera systems. Unlike traditional manual or semiautomatic photogrammetric methods, the presented approach aims to reduce human intervention while maintaining geometric accuracy. The workflow includes camera parameter estimation, geometric correction of raw images, automatic extraction of relevant image segments, spatial alignment of image strips, and final mosaicking to generate a complete facade texture. Experimental results show that the method is capable of handling images with significant geometric distortions and provides reliable results for urban modeling applications.
