为了提高播种导航的定位效率和精度,需要引入较为先进的图像处理算法,而对播种机视觉导航传感器采集得到的实时视频进行深度图像预测及处理则是其中的关键技术.基于色度图像分割的方法可以有效地处理深度图像的预测噪声,而采用专用多核DSP 系统可以完成深度图像的处理及优化.为了验证方案的可行性,以无人驾驶播种机导航系统为例,对导航所需时间以及导航精度进行了测试.测试结果表明:采用深度图像处理技术可以得到较高的导航精度;采用多核处理器的效率要明显高于单核处理器,处理精度也有所提升,可以满足无人驾驶播种机导航系统的设计需求.%In order to improve the efficiency and accuracy of navigation, it is necessary to introduce more advanced image processing algorithm. And the key technology is advanced image prediction and processing for the real-time video captured by seeder vision navigation sensor. The prediction noise of advanced image can be effectively processed by color image segmentation method, and the processing and optimization of advanced image can be completed by special multi-core DSP system. In order to verify the feasibility of the scheme, the navigation time and precision of the unmanned seeder navigation system are tested. The results show that higher navigation accuracy can be achieved by the image advanced processing technology, the efficiency of multi-core processor is obviously higher than that of the single-core processor, and the accuracy of the system has also been improved. The project can meet the design requirements of the navigation system of the unmanned seeder.
展开▼
