Generation of a three-dimensional model of an object based on its photographic images
The creation of drawings and object models constitutes an essential component of a wide range of scientific, engineering, and even everyday tasks. Consider, for instance, a situation in which a new technological device must be designed to conduct scientific experiments. In such a case, it is necessary to develop a detailed model of the object to support precise calculations of various characteristics. Another example can be found in civil engineering, where, for instance, the strength analysis of a bridge requires a complete model with all geometric dimensions and material properties to correctly estimate the maximum permissible load. Even in everyday applications, three-dimensional (3D) models are increasingly used — for example, in the design and 3D printing of small household items. All these cases require specialized software that allows for working with 3D models. One of the fundamental concepts of photogrammetry is central projection. When photographing an object, its image is always subject to distortion, regardless of viewpoint, making it impossible to perform accurate measurements directly from raw photographs. Therefore, before images can be used as cartographic or metric materials, they must undergo specific photogrammetric processing. These distortions arise because the recording device is typically located at a single position, and the image is therefore formed by central projection. This work addresses the creation of a 3D model of an object using ordinary photographs, with the aim of making the process accessible to non-specialist users — that is, without requiring deep knowledge of photogrammetry or expensive equipment. The study considers the principles of central projection, its key points, and lines. Based on this theoretical foundation, corresponding points between photographs were identified, followed by the generation of a sparse point cloud. The cloud was subsequently densified, and a three-dimensional model of the object was constructed, with the final step involving texture mapping. The proposed approach enables relatively fast 3D reconstruction of small objects without the use of costly instruments. Such a method can be effectively applied to create three-dimensional models of laboratory instruments and devices, which can in turn be used in remote learning environments to simulate and teach practical work with real equipment. Keywords: 3D model, photographic surveying, point cloud, central projection, search for common points between photographs.
Read more