This study presents comprehensive studies on practical means to predict aerial-image motion blurs due to the flying vehicle dynamics, including flying altitude; cruising speed; and angular velocities and finally installed camera characteristics; such as, frame rate and image size. The resulting predictions of blur values are in-turn used to generate blurry images to be fed as input data for later use in a de-blurring-in-the-loop of a Mono-simultaneous localization and mapping system. The whole process is coordinated by means of an integrated aerial virtual environment. The integrated aerial virtual environment consists of a three-dimensional graphical engine which could communicate with a full six-degrees of freedom aircraft dynamic simulator to precisely generate trajectories of the flying vehicle. Any accumulated real-time image taken by the camera installed on the aircraft during its mission considers the relevant vehicle states for necessary adjustments. We practically exploit Cþþ, High-Level Shader Language, as well as JSBSim to expand the components of the package. Extensive case-studies show that the developed approach is quite effective in controlled environments.
This paper offers an algorithm for enhancement of positioning accuracy of a quad-rotor flying robot, based on jerk and jounce of motion. The suggested method utilises the first and second numerical derivatives of the vehicle's acceleration and augments the mathematical model in the estimation process. For this purpose, the Kalman Filter (KF) is implemented for integration of a Strapdown Inertial Navigation System (SINS) and Global Navigation Satellite System (GNSS). The required data are collected from a low-cost/quality Micro Electromechanical Sensors (MEMS) during an assisted flight. For increasing the precision and accuracy of the collected data, all instruments including accelerometers, gyroscopes and magnetometers are calibrated before the experiments. Moreover, to reduce and limit the measurement noises of the MEMS sensor, a low-pass filter is applied; this is while sensors in the autopilot are affected by high levels of noise and drift, which makes them inappropriate for accurate positioning. The experimental results exhibit an improvement in positioning and altitude sensing through augmentation of the loosely coupled SINS/GNSS navigation method.
scite is a Brooklyn-based organization that helps researchers better discover and understand research articles through Smart Citations–citations that display the context of the citation and describe whether the article provides supporting or contrasting evidence. scite is used by students and researchers from around the world and is funded in part by the National Science Foundation and the National Institute on Drug Abuse of the National Institutes of Health.