MOBILE-CLOUD SYSTEM FOR SOLVING PHOTOGRAMMETRY TASKS IN INDUSTRY

Abstract

With the development of the capabilities of mobile devices and the increase in the availability of wireless communication, the possibilities of building industrial automation systems have significantly expanded. The quality of digital photography obtained with a smartphone camera makes it possible to build mobile systems based on computer vision: for example, photogrammetry systems. There are several factors to consider. The first factor is that the tasks of processing digital photography for industrial purposes remain resource-intensive and cannot be fully implemented only on the basis of a mobile device. Therefore, it is required to transfer the execution environment for resource-intensive tasks to third-party computing power available on demand. The second factor is the stability and bandwidth of the communication channel - mobile devices are usually needed in remote locations where the deployment of desktop computers is not possible. Therefore, using a smartphone only as a camera is not always justified, since the transfer of an unprocessed image may take a long time or even be impossible. The third factor hindering the widespread use of mobile devices in solving photogrammetric problems is the variability and constant emergence of new methods of image processing and analysis. It is necessary to centrally create and replenish libraries of such modules. Thus, the creation of mobile photogrammetric measuringsystems requires combining the computing power of cloud services and the mobility of smartphones. The article proposes a method for constructing photogrammetric measuring systems based on mobile cloud computing, which provides a dynamic balance of the computational load on the nodes of the system, as well as the variability of functionality on mobile devices of users