夏文泽, 韩绍坤, 曹京亚, 王亮, 翟倩. 激光雷达距离估计技术[J]. 红外与激光工程, 2016, 45(9): 906005-0906005(6). DOI: 10.3788/IRLA201645.0906005
引用本文: 夏文泽, 韩绍坤, 曹京亚, 王亮, 翟倩. 激光雷达距离估计技术[J]. 红外与激光工程, 2016, 45(9): 906005-0906005(6). DOI: 10.3788/IRLA201645.0906005
Xia Wenze, Han Shaokun, Cao Jingya, Wang Liang, Zhai Qian. Ladar range estimation technology[J]. Infrared and Laser Engineering, 2016, 45(9): 906005-0906005(6). DOI: 10.3788/IRLA201645.0906005
Citation: Xia Wenze, Han Shaokun, Cao Jingya, Wang Liang, Zhai Qian. Ladar range estimation technology[J]. Infrared and Laser Engineering, 2016, 45(9): 906005-0906005(6). DOI: 10.3788/IRLA201645.0906005

激光雷达距离估计技术

Ladar range estimation technology

  • 摘要: 距离估计技术一直是激光雷达技术中的核心技术,对整个激光雷达系统性能的优劣性起到决定性作用。提出了一种基于人工神经网络的距离估计算法,使用的人工神经网络由两层网络构成。与交叉相关距离估计算法的对比实验证明:基于人工神经网络的距离估计算法在实际应用中有更高的距离估计精度;对强度像进行了仿真实验,实验结果证明:基于人工神经网络的距离估计算法相比极值判别法能够获得更精准的回波强度信息;对神经网络隐含层中神经元个数与距离估计精度之间的关系进行了实验分析,实验结果证明:过多或过少的神经元个数均会降低距离估计精度。

     

    Abstract: Range estimation technology, which plays a decisive role on the properties of the laser radar system, is always the core technology of laser radar technology. A range estimation technology based on Artificial Neural Network which contains two layers of network was proposed. The contrast experiment with the Cross-correlation range estimation technology demonstrates this technology can provide more accurate range estimation result. A simulation experiment on intensity image was conducted, and experiment result show this technology can provide more accurate intensity data than the peak judgment technology. The correlation between the neuron amount of hidden layer and range estimation accuracy was also analyzed, and the result shows too many or too few neurons will decrease the range estimation accuracy.

     

/

返回文章
返回