iarjetIARJETIARJET2708-5163https://doi.org/10.47310/iarjet.2021.v02i02.020Methods Development for Photogrammetry Results ImprovementsBorodavkaYevheniiKharchenkoOleksandrProfessor, Doctor of Technical Sciences, Kyiv National University of Construction and Architecture, KyivGraduate Student of the Department of Information Technologies, Kyiv National University of Construction and Architecture, KyivModern photogrammetry has potential in many areas of human activity from construction to meteorology. Over the past five years, technologies and methods have been developed in this discipline that challenge traditional views and approaches in these fields. In particular, the introduction of structures on the methodology of movement has led to a significant increase in the use of photogrammetry in geological and engineering-geological practice. Achievements have been achieved mainly through social, political, environmental and technological changes around the world. Human mobility has increased significantly due to population growth, climate change and globalization. Innovations in photogrammetry have also been strongly influenced by the development of information and communication technologies, robotics and computer vision. With the use of remote sensing and radar techniques, the ability to collect, analyze and integrate data has greatly increased even for reflective surfaces such as metal or glass and uniform textures such as snow or ice. The availability of new high-resolution digital cameras and photogrammetry software has led to a gradual improvement in the quality of the surface data of the objects to be collected. Creating photogrammetric 3D models is now much faster and easier thanks to the use of motion structures but the use of ground-based control points to scale the model is still required. The paper considers the shortcomings of photogrammetry methods, quality problems and the need for additional processing of the final result. The reason for these shortcomings is that when using standard methods of photogrammetry, it is impossible for the software to evaluate the desired result and take any measures to improve it. After analyzing the general stages of the photogrammetry process, it is proposed to use computer vision and in particular the technique of object recognition using models based on machine or deep learning, which will help identify "places of increased attention" and remove useless data.