机车仿真器视景系统中的图象处理

Image Processing in the Visual System of Locomotive Simulator

为了满足视景仿真中空间分辨率的要求，将传统的几何变换和小波变换相结合，提出了一种改进的图象放大算法。该算法能够在不丢失高频信息的条件下放大图象，使得放大后的图象画面清晰，细节丰富。为了叠加信号灯，提出了基于模式识别的区域连接增长算法，该算法能够自动跟踪信号灯的位置和大小。

In order to improve the spatial resolution of visual simulation, An immproved algorithm which cooperates the wavelet transform with the traditional geometry is proposed. Itcan enlarge images without loss of high frequency information, and make the enlarged imageclearer and richer. A region join growing algorithm based on pattern recognition which cantrack the location and size of signal panels automatically is put forward to overlay the signals.