Image Rendering Research of Image Guidance Simulator
|School||National University of Defense Science and Technology|
|Course||Control Science and Engineering|
|Keywords||Virtual reality technology Image stitching technology SIFT features Brightness correction Spherical panorama Visual Simulation System|
Image-guided simulator is used for testing image-guided weapons guidance systems, as well as training in the use of image-guided weapons warfighters typical equipment, including how to achieve realistic image generation (analog seeker acquisition) target area image guided simulation is one of the keys. Image generation method of generating virtual modeling and image stitching, image stitching more realistic and access to a wider range of applications. Therefore, the image of the image-guided simulator Health research background, research carried out based on SIFT (Scale Invariant Feature Transform) feature points in image stitching method, spherical panorama generation and image guided simulator visual simulation systems. Firstly, image mosaic based on feature points matching details SIFT feature extraction, image registration algorithm based on RANSAC (random sample consensus) specific image transformation matrix calculation steps, and put forward a kinds of improved the Hat function fusion algorithm fusion Hat function before adding global correction link. Convert all the RGB image to the HSV space, respectively, to adjust the image of the hue, saturation, luminance, of each of all the images of the hue, saturation, brightness mean value as the reference standard, to adjust them to a more balanced range, restore the image of the HSV space and then adjusted to RGB format, and then the Hat function method adjusted image fusion, stitching results are not due to the sampling of the image color difference ghosting or fuzzy algorithm and simulation experiments effectiveness. Secondly, in order to improve the local quality of image stitching explore spherical panorama generation method. Spherical model, spherical coordinate plane coordinates corresponding relationship derived in detail. The overall optimization methods modified to avoid the cumulative error in the panorama generated transformation matrix for image registration. When image stitching, will the Hat function method improved extended into a spherical form, to achieve a seamless integration of the panorama. On the basis of the above study, image-guided simulator visual simulation. Additional restrictions nearest neighbor node search method and appropriate visual image drawing method, real-time simulation of the visual image of the long-range missiles into the seeker for terminal guidance phase. Can use the system to test the control system, pilot training, and to reduce the test cost and risk. Generate real-time seeker field of view image has important applications for imaging-guided study.