ENRI papers

The evaluation of fusion image based on the experimental enhanced vision system

Yasuto SumiyaCShigeru Ozeki and Masayuki Shirakawa

Abstract:
EVSiEnhanced Vision Systemjand SVSiSynthesized Vision Systemjare known to be effective for pilots to improve situational awareness. An experimental EVS was developed utilizing image fusion techniques for our evaluation. The system was developed and evaluated in two phases. In the first phase, the system obtained the GPS positioning data, attitude data from Gyro sensor and the actual motion picture of scenery on the ground or the ENRI's experimental aircraft to create fusion images. In the second phase, the system integrated the three dimensionali3Djimage using the afore-mentioned data and digital map database of the Geographical Survey Institute. The graphical system is used to process the 3D image for expressing the surrounding image from sensor's position. The system carries out a fusion of the 3D image and the actual motion pictures. Since the system is used on the ground, that is, a post-experimental analysis, the flight crews cannot see the fusion images while flying.

Two methods are examined in the simulation of the fusion system for these images. One is a method that the artificial image is overlapped with the actual images of motion picture by a computer to display on its screen. The other is a method that the HUDiHead Up Displayjis used to overlap both images and to provide the view for its observer. This paper discusses the difference of two methods as fusion systems. The method with HUD can provide the higher update rate for the animation, for example, with several frames per second or more. However, the accurate adjustment of the observer's position is required.

Several smoothing and prediction methods for measured position and attitude data are also discussed to synchronize the artificial images to actual images. This paper presents the results of the comparison.

The in-house pilot evaluation of the experimental system is carried out for the fusion system with HUD. The evaluation items are scrutinized for the frame rate, the contrast and the color of the 3D image. It also presents the results of the evaluation. It was evaluated that the maximum frame rate of more than 12 is suitable. The best color of the 3D image was a complementary color of the background picture's color.