机译:基于卷积神经网络和归一化互相关的导航地磁匹配方法
Korea Adv Inst Sci & Technol Dept Aerosp Engn 291 Daehak Ro Daejeon South Korea;
Hanwha Corp 10 Yuseong Daero 1366 Beon Gil Daejeon South Korea;
geomagnetism; learning (artificial intelligence); navigation; image matching; convolutional neural nets; correlation methods; image classification; numerical analysis; geomagnetic matching; convolutional neural network; normalised cross-correlation; geomagnetic information; neural networks; reference map; map positions; symmetric convolutional NNs; NCC; symmetric CNNs; matching position; geomagnetic navigation area; geomagnetic measurements; mean matching rate; worst geomagnetic profile; initial position estimation; convergence time; position estimation error; measurement vector mapping; numerical simulation;
机译:基于卷积神经网络的声纳图像匹配自主水下航行器导航
机译:基于卷积神经网络的声纳图像匹配自主水下航行器导航
机译:使用基于复发绘图的卷积神经网络改善室内地磁场指纹识别
机译:基于卷积神经网络的声纳图像匹配自动水下车辆导航
机译:基于卷积神经网络的人体个性化3D建模方法及其分类
机译:基于速度基于区域的卷积神经网络的正畸患者在正畸患者中检测早期学习的方法
机译:基于卷积神经网络的导航地磁匹配方法,归一化互相关