为提高无人机避障的灵活性和可靠性,提出了一种基于LGMD(Lobula Giant Movement Detector)的无人机避障方法,通过将视场分割为上、下、左、右4个方位,形成4个方位竞争的LGMD(C-LGMD),并利用Matlab软件进行算法实现和视频仿真分析,最后...为提高无人机避障的灵活性和可靠性,提出了一种基于LGMD(Lobula Giant Movement Detector)的无人机避障方法,通过将视场分割为上、下、左、右4个方位,形成4个方位竞争的LGMD(C-LGMD),并利用Matlab软件进行算法实现和视频仿真分析,最后将算法移植到无人机硬件系统,开展悬停测试和实时飞行实验研究。由视频仿真分析和悬停测试结果表明,该算法能有效分辨来自不同方位的障碍物,具有较好的避障性能和鲁棒性;在实时飞行测试中,无人机在室内环境中可实现三维空间有效避障,验证了该算法的可靠性。研究结果为进一步探索无人机高效、可靠避障提供参考依据。展开更多
公共场所中的人群突发局部聚集常是异常事件发生的先兆,由于其随机性强,前兆特征不明显,现有的传统计算机视觉技术较难对其有效检测。基于蝗虫视觉系统的神经结构特性与小叶巨型运动检测器(lobula giant movement detector,LGMD)危险感...公共场所中的人群突发局部聚集常是异常事件发生的先兆,由于其随机性强,前兆特征不明显,现有的传统计算机视觉技术较难对其有效检测。基于蝗虫视觉系统的神经结构特性与小叶巨型运动检测器(lobula giant movement detector,LGMD)危险感知机理,提出一种人群突发局部聚集行为检测的LGMD改进型神经网络模型。该模型感知人群活动在视野域中引发的视觉信号,基于哺乳动物视网膜视觉信号处理机制整合视觉运动线索,借助LGMD神经元危险感知机理构建尖峰阈值机制调谐神经网络输出,以感知人群活动中的突发聚集行为。不同场景下的人群活动视频实验结果表明,提出的神经网络能有效检测视野域中人群突发局部聚集行为并对其预警。该文涉及生物视神经机理启发的人群活动动态视觉信息加工处理,可为智能视频监控中的人群活动检测与行为分析提供新思想、新方法。展开更多
在移动机器人、自动驾驶、视频监控等应用领域,复杂的动态场景中,对于物体深度运动及方向检测一直是计算机视觉技术的难点。自然界中,昆虫在飞行过程中利用复眼视觉检测高度变化且视觉杂乱环境中的深度运动物体(或称为目标),是学习运动...在移动机器人、自动驾驶、视频监控等应用领域,复杂的动态场景中,对于物体深度运动及方向检测一直是计算机视觉技术的难点。自然界中,昆虫在飞行过程中利用复眼视觉检测高度变化且视觉杂乱环境中的深度运动物体(或称为目标),是学习运动感知策略的良好范例。受飞行昆虫复眼视觉功能优势的启发,本文采用物体运动的缩放变量计算与基于LGMD (lobula giant movement detector)的改进型碰撞检测模型相结合的仿生策略,提出一种基于LGMD神经元建模的物体深度运动方向估计方法(简称LGMD-ED),通过对PPT合成动画视频和拍摄真实场景的样本视频进行仿真实验和测试,验证了本文所提新方法对于检测与估计物体远离和靠近二种典型深度运动方向的有效性。展开更多
LGMD type 21, caused by mutations in the fukutin-related protein, is a common form of LGMD. The phenotype resembles Duchenne/Becker muscular dystrophy. A point mutation, L276I has been found in all patients with LGMD2...LGMD type 21, caused by mutations in the fukutin-related protein, is a common form of LGMD. The phenotype resembles Duchenne/Becker muscular dystrophy. A point mutation, L276I has been found in all patients with LGMD2I studied so far. The authors screened for this mutation in 102 sporadic cases of Duchenne/Becker mutation-negative patients and found 13 patients with LGMD2I.展开更多
Objective: To determine the phenotype variability associated with the specific C-terminal M-line titin mutation known to cause autosomal dominant distal myo pathy, tibial muscular dystrophy (TMD; MIM 600334), and limb...Objective: To determine the phenotype variability associated with the specific C-terminal M-line titin mutation known to cause autosomal dominant distal myo pathy, tibial muscular dystrophy (TMD; MIM 600334), and limb girdle muscular dys trophy 2J (LGMD2J). Methods: Three hundred eighty-six individuals were genotype d for the Finnish founder mutation in titin (FINmaj) causing TMD/LGMD2J. Results : Two hundred seven patients were heterozygous for the mutation. Among these pat ients, 189 (91%) had a more common phenotype compatible with the classic descri ption of TMD. However, 18(9%) had unusual phenotypes such as proximal leg or po sterior lower leg muscle weakness and atrophy even at onset.Four patients were c onfirmed homozygotes representing the LGMD2J phenotype. These homozygotes were h alf of the eight LGMD patients previously described in the original large consan guineous kindred. Conclusions: Large variability of phenotypic expression caused by just one mutation, the Finnish FINmaj, suggests that no certain phenotype of myopathy/dystrophy can be excluded from being caused by mutated titin. Yet unkn own homozygous or compound heterozygous titin mutations without phenotype in the heterozygote carriers may be responsible for undetermined recessive MD and LGMD .展开更多
文摘为提高无人机避障的灵活性和可靠性,提出了一种基于LGMD(Lobula Giant Movement Detector)的无人机避障方法,通过将视场分割为上、下、左、右4个方位,形成4个方位竞争的LGMD(C-LGMD),并利用Matlab软件进行算法实现和视频仿真分析,最后将算法移植到无人机硬件系统,开展悬停测试和实时飞行实验研究。由视频仿真分析和悬停测试结果表明,该算法能有效分辨来自不同方位的障碍物,具有较好的避障性能和鲁棒性;在实时飞行测试中,无人机在室内环境中可实现三维空间有效避障,验证了该算法的可靠性。研究结果为进一步探索无人机高效、可靠避障提供参考依据。
文摘公共场所中的人群突发局部聚集常是异常事件发生的先兆,由于其随机性强,前兆特征不明显,现有的传统计算机视觉技术较难对其有效检测。基于蝗虫视觉系统的神经结构特性与小叶巨型运动检测器(lobula giant movement detector,LGMD)危险感知机理,提出一种人群突发局部聚集行为检测的LGMD改进型神经网络模型。该模型感知人群活动在视野域中引发的视觉信号,基于哺乳动物视网膜视觉信号处理机制整合视觉运动线索,借助LGMD神经元危险感知机理构建尖峰阈值机制调谐神经网络输出,以感知人群活动中的突发聚集行为。不同场景下的人群活动视频实验结果表明,提出的神经网络能有效检测视野域中人群突发局部聚集行为并对其预警。该文涉及生物视神经机理启发的人群活动动态视觉信息加工处理,可为智能视频监控中的人群活动检测与行为分析提供新思想、新方法。
文摘在移动机器人、自动驾驶、视频监控等应用领域,复杂的动态场景中,对于物体深度运动及方向检测一直是计算机视觉技术的难点。自然界中,昆虫在飞行过程中利用复眼视觉检测高度变化且视觉杂乱环境中的深度运动物体(或称为目标),是学习运动感知策略的良好范例。受飞行昆虫复眼视觉功能优势的启发,本文采用物体运动的缩放变量计算与基于LGMD (lobula giant movement detector)的改进型碰撞检测模型相结合的仿生策略,提出一种基于LGMD神经元建模的物体深度运动方向估计方法(简称LGMD-ED),通过对PPT合成动画视频和拍摄真实场景的样本视频进行仿真实验和测试,验证了本文所提新方法对于检测与估计物体远离和靠近二种典型深度运动方向的有效性。
文摘LGMD type 21, caused by mutations in the fukutin-related protein, is a common form of LGMD. The phenotype resembles Duchenne/Becker muscular dystrophy. A point mutation, L276I has been found in all patients with LGMD2I studied so far. The authors screened for this mutation in 102 sporadic cases of Duchenne/Becker mutation-negative patients and found 13 patients with LGMD2I.
文摘Objective: To determine the phenotype variability associated with the specific C-terminal M-line titin mutation known to cause autosomal dominant distal myo pathy, tibial muscular dystrophy (TMD; MIM 600334), and limb girdle muscular dys trophy 2J (LGMD2J). Methods: Three hundred eighty-six individuals were genotype d for the Finnish founder mutation in titin (FINmaj) causing TMD/LGMD2J. Results : Two hundred seven patients were heterozygous for the mutation. Among these pat ients, 189 (91%) had a more common phenotype compatible with the classic descri ption of TMD. However, 18(9%) had unusual phenotypes such as proximal leg or po sterior lower leg muscle weakness and atrophy even at onset.Four patients were c onfirmed homozygotes representing the LGMD2J phenotype. These homozygotes were h alf of the eight LGMD patients previously described in the original large consan guineous kindred. Conclusions: Large variability of phenotypic expression caused by just one mutation, the Finnish FINmaj, suggests that no certain phenotype of myopathy/dystrophy can be excluded from being caused by mutated titin. Yet unkn own homozygous or compound heterozygous titin mutations without phenotype in the heterozygote carriers may be responsible for undetermined recessive MD and LGMD .
