Assessing a 35mm Fixed-Lens Sony Alpha-5100 Intrinsic Parameters Prior to, During, and Post UAV Flight Mission

Abstract

A fixed focal length lens (FFL) camera with on-adjustable focal length is common companions for conducting aerial photography using unmanned aerial vehicles (UAVs) due to its superiority on optical quality and wider maximum aperture, lighter weight and smaller sizes. A wide-angle 35mm FFL Sony a5100 camera had been used extensively in our recent aerial photography campaign using UAV. Since this off-the-self digital camera is categorized into a non-metric one, a stability performance issue in terms of intrinsic parameters raises a considerably attention, particularly on variations of the lens principal distance and principal point’s position relative to the camera’s CCD/CMOS sensor caused by the engine and other vibrations during flight data acquisitions. A series of calibration bundle adjustment was conducted to determine variations in the principal distances and principal point coordinates before commencing, during, and after accomplishment of the flight missions. This paper demonstrates the computation of the parameters and presents the resulting parameters for three different epochs. It reveals that there are distinct discrepancies of the principal distances and principal point coordinates prior to, during, and after the mission, that peaked around 1.2mm for the principal distance, as well as around 0.4mm and 1.3mm along the x-axis and the y-axis of the principal point coordinates respectively. In contrast, the lens distortions parameters show practically no perturbations in terms of radial, decentering, and affinity distortion terms during the experiments.

Cadastral Surveys with Non-metric Camera Using Uav: A Feasibility Study

“…Due to Conventional flight configurations for aerial image acquisitions using small format cameras are uneconomic for the full provision of ground control points since each photograph covers a small area of ground [11], calibration are performed without necessarily measured GCPs on the field. Flight configurations are followed convergent network geometry to facilitate block invariant self-calibrating bundle adjustment model to calibrate the camera intrinsic parameters [12] as well as determine object space coordinates of pre-marks' locations [13,14]. The followings are general forms of the method in the form of Cayley matrix [15,16]:…”

“…In the self-calibrating bundle adjustment process, provision of a fixed GCPs array on the calibration test field has no or little impact on the quality of the adjusted parameters. However, it requires the pre-mark's array be well distributed in three dimensions [13,14] and it is necessary for a highly convergent imaging geometry for a recovery of the principal distance [15,16] (i.e. calibrated focal length).…”

Potential Use of UAV-Based Mapping System to Accelerate the Production of Parcel Boundary Map in Indonesia

“…Burgeoning Unmanned Aerial Vehicle (UAV) which can carry a consumer grade camera GEODETA 2019 [2] open the horizon of such a solution due to a very dense ground sampling distance (GSD) provided and particularly it gives a cost effective solution of mapping activities in terms of portability of the platform [3], quick and cheap mapping production line [4,5], and the images geometric accuracies are readily improved by using well calibrated camera [6,7].…”

Geometric Accuracy Assessments of Orthophoto Production from UAV Aerial Images

“…The use of consumer grade cameras mounted on UAV coupled with a so called structure from motion (SfM) method is widely applied for orthophoto productions [26,27]. SfM automates a computation of Interior Orientation (IO) parameter, and lens distortion of the camera [6,8], and Exterior Orientation (EO) parameters of all images without the need of known 3D coordinates of control points. It also incorporates Multi-View Stereopsis (MVS) method to derive dense point clouds from overlapping aerial images acquired from multiple altitudes, locations and view angles [28,29].…”

“…The transformation accuracy can be further improved by including lens distortion model (Δx, Δy) into the collinearity equation of (1). This is the case since consumer grade cameras are utilized for aerial photographing, as their lenses are of lower quality than those of metric cameras [6,8]. The coefficients implemented are correcting radial distortions, decentring distortions, as well as affinities.…”