摘要
提出了一种全景视觉泊车辅助装置及其生成全景视觉图像的方法,通过全景视觉泊车辅助装置合成为一个虚拟的鸟瞰全景图像,显示在屏幕上,能够实时提供驾驶员泊车所需要的汽车全景图像,消除了车四周的视觉盲区,在此基础上添加偏航模块,利用标尺线对车进行实时定位,以帮助驾驶员更加精确的泊车。生成全景视觉图像的算法主要包含去鱼眼失真、平面投影映射和图像合成。通过DSP处理器根据预先存放在FLASH中的虚拟全景图像查找列表进行寻址操作得到虚拟的全景图像。
The paper puts forward a parking auxiliary device based on panoramic visual image and a method of generating panorama of visual images. A virtual panoramic bird's eye view can supply realtime panoramic images which displayed on the screen when the driver parking cars. The device can eliminate the blind spots around the car. On the basis of above, the yaw module is added, and the vehicle is positioned on the base of the scale line to make the system more safety and more accurate. The algorithm of generating panorama vision mainly involve adjusting fisheye image distortion, planar perspective projection mapping and image synthesis. The virtual panoramic image which can be get by addressing the image list on the DSP processor is stored in the FLASH in advance.
出处
《科技通报》
北大核心
2017年第11期170-173,共4页
Bulletin of Science and Technology
关键词
全景视觉
鸟瞰
投影映射
图像合成
panoramic view
bird eye
perspective projection
image synthesized
