PHOTOGRAMMETRIC SURVEYING AND 3D MODELING USING UNMANNED AERIAL VEHICLE ( DRONES ) ----- PART 1

DRONEWORK

Photogrammetry, UAVs, Flight planning, DEM generation, Orthophoto, 3-D Modelling, UAV image classification , Illumination conditions for different scenes .

Unmanned Aerial Vehicle (UAV), generally known as a drone are nowadays a worthy source of data for 2D & 3D mapping, quick inspection & continuous monitoring, and for many other applications. Their ability in quick deployment, manoeuvre, and fast data collection at a very high spatial and temporal resolution is unprecedented. In this study we have attempted to understand the usability of UAV for planimetric and topographic mapping. We have tried to capture various man-made and natural objects using different viewing geometry, flight plans and in various acquisition modes. Initially, the captured images were aligned, and first level point clouds were created using photogrammetric techniques which depend upon camera position (using GPS and viewing geometry). These 1st-level points were densified at the 2nd level and then 3D-mesh were created. Finally, ortho-mosaic covering the region of capture was derived, and 3D-stereoscopic image of the study area was generated which helps in visualising the captured area interactively. Planimetric measurements (such as height, surface area) were carried out using markers on the 3D-tiled model. Volumetric analysis was also carried out on the 3D-solid mesh data using free-style selection. Overall 10 different results were generated: 3D-solid mesh, wireframe, textured model, tiled model, DEM, DSM, contour lines, stereoscopic view, ortho-mosaic and video using individual ortho-rectified images. We have utilise various existing softwares (in demo mode) such as Agisoft Photoscan, Pix4D Capture, Pix4D Mapper, and Drone Deploy for various purposes. We have validated the UAV derived measurements with the ground measurements and the error varied within few centimetres which is unimaginable (in terms of accuracy, cost and time efficiency) with the current existing satellite based or flight based technologies. Classification of the images were also carried out, Image classification refers to the task of extracting information classes from a multiband raster image. Both supervised and unsupervised classification were performed on the orthomosaic data of the UAV image. With the ArcGIS Spatial Analyst extension, there is a full suite of tools in the Multivariate toolset to perform supervised and unsupervised classification .The classification process is a multi-step workflow. Principal component analysis was also done on the data it has been used for visualization of complex data and developed to capture as much of the variation in data as possible. After the successful classification of the orthomosaic data the accuracy assessment of classified data were performed .In the context of information extraction by image analysis, accuracy “measures the agreement between a standard assumed to be correct and a classified image of unknown quality.Classification error occurs when a pixel (or feature) belonging to one category is assigned to another category. Errors of omission occur when a feature is left out of the category being evaluated; errors of commission occur when a feature is incorrectly included in the category being evaluated so that the project can be completed with more accuracy. 
    

                        

UAV is an unmanned-winged system, functioned remotely by a human operator or independently by an on-board system. UAV is also  defined as a "powered, aerial vehicle that does not carry a human operator, uses aerodynamic forces to provide vehicle lift, can fly autonomously or be piloted remotely, can be expendable or recoverable, and can carry a lethal or nonlethal payload" (http://TheFreeDictionary.com.).The term UAV is generally used in the branch of science that deals with the collection, analysis, and interpretation of data relating to the earth's surface. Based on their size, weight, endurance & range and flying altitude, it is also terms as Remotely Piloted Vehicle (RPV), Remotely Operated Aircraft (ROA), Remote Controlled (RC) Helicopter, Unmanned Vehicle Systems (UVS) etc.(Remondino et al., 2011). Currently existing satellites and manned vehicles are the main source of obtaining spatial data, but they are unable to collect data in real time, as well as they are too expensive. Therefore, we need a cheaper, more suitable and real-time approach to get spatial data. Compared with spaceborne remote sensing, UAV remote sensing has a feature of higher temporal resolution, and can also avoid the effect of cloud; relative to airborne remote sensing(Barazzetti et al., 2014).UAV based Remote Sensing is the new addition for significant mapping, real-time survey and monitoring activities for numerous applications. As compared to existing piloted aerial system, UAV is considerably cheaper to use, easy to handle and manoeuvre (Fraseret al., 2015). It can be used (take-offs and landings) even in inaccessible areas and has a low noise operation as compare to other devices. With the help of UAVs we can collect high spatial resolution, multi-perspective, multi sensor imagery over a study site(Fritz et al., 2017).A typical image-based field surveying with UAV systems require a flight or mission planning, GCPs measurement,image acquisition, camera calibration, image orientation and image processing for 3D information extraction (Grenzdörffer et al., 2011).