基于四元数显著图和PCNN空洞滤波的足球检测

注意力选择和先验知识可有效的用于目标检测与跟踪。在基于注意力选择目标跟踪模型的基础上,引入了PCNN（Pulse Coupled Neural Network）空洞滤波及足球颜色的先验知识。一方面,针对有些情况下足球模糊不清,采用PCNN空洞滤波检测足球的连通性;另一方面,在注意力选择模型中引入了足球颜色的先验知识,进一步提高检测性能。首先提取球场区域,然后对由足球颜色先验知识、色差和亮度产生的四元数显著图进行PCNN空洞滤波,很多情况下可由此直接检测到足球。如果至此未检测到足球,继续利用四元数显著图生成感兴趣区域,并用足球的面积、圆形度和离心率等特征进一步检测目标,同时采用卡尔曼滤波器预测足球的位置作为补充检测。仿真结果显示,与Dynamic Kalman Filter with Velocity Control^[4]和Real Time Ball Detection Framework^[5]两种方法相比,检测成功率分别提高了11．5％和15．8％。     

Marginalized particle filter for spacecraft attitude estimation from vector measurements

An algorithm based on the marginalized particle filters （MPF） is given in details in this paper to solve the spacecraft attitude estimation problem： attitude and gyro bias estimation using the biased gyro and vector observations. In this algorithm, by marginalizing out the state appearing linearly in the spacecraft model, the Kalman filter is associated with each particle in order to reduce the size of the state space and computational burden. The distribution of attitude vector is approximated by a set of particles and estimated using particle filter, while the estimation of gyro bias is obtained for each one of the attitude particles by applying the Kalman filter. The efficiency of this modified MPF estimator is verified through numerical simulation of a fully actuated rigid body. For comparison, unscented Kalman filter （UKF） is also used to gauge the performance of MPE The results presented in this paper clearly derfionstrate that the MPF is superior to UKF in coping with the nonlinear model.     

Cubic helical splines with Frenet-frame continuity

We present a scheme of interpolating a sequence of points p₀,…,pn in R³ by a piecewise-cubic-helical spline with continuous Frenet frame. Each cubic helical segment rk(ξ) connecting pk−1 and pk is constructed such that its end tangents are symmetric with respect to the displacement vector Δpk=pk−pk−1. The existence condition of such a segment rk(ξ) is formulated in terms of the configuration involving Δpk and the Frenet frame of rk(ξ) at pk−1. A Frenet-frame-continuous spline is obtained by matching the Frenet frames of sequentially constructed segments at their juncture points.     

Dual tree fractional quaternion wavelet transform for disparity estimation

This paper proposes a novel phase based approach for computing disparity as the optical flow from the given pair of consecutive images. A new dual tree fractional quaternion wavelet transform (FrQWT) is proposed by defining the 2D Fourier spectrum upto a single quadrant. In the proposed FrQWT, each quaternion wavelet consists of a real part (a real DWT wavelet) and three imaginary parts that are organized according to the quaternion algebra. First two FrQWT phases encode the shifts of image features in the absolute horizontal and vertical coordinate system, while the third phase has the texture information. The FrQWT allowed a multi-scale framework for calculating and adjusting local disparities and executing phase unwrapping from coarse to fine scales with linear computational efficiency.     

A structure-preserving Jacobi algorithm for quaternion Hermitian eigenvalue problems

A new real structure-preserving Jacobi algorithm is proposed for solving the eigenvalue problem of quaternion Hermitian matrix. By employing the generalized JRS-symplectic Jacobi rotations, the new quaternion Jacobi algorithm can preserve the symmetry and JRS-symmetry of the real counterpart of quaternion Hermitian matrix. Moreover, the proposed algorithm only includes real operations without dimension-expanding and is generally superior to the state-of-the-art algorithm. Numerical experiments are reported to indicate its efficiency and accuracy.     

六自由度机器人圆弧平滑运动轨迹规划

通过空间3点,求出空间圆弧的参数方程;根据机器人运动学转换矩阵,将空间圆弧参数方程转换到圆心所在平面内,利用7次多项式方法进行规划.通过四元数球面线性插值,对机器人姿态进行规划,得到平滑圆弧运动.     

基于四元同质微分的彩色光流估计

该文通过将颜色3维矢量表示为一个4元数(quaternion)形式的超复数,将彩色图像序列映射为3维时-空空间上定义的4元超复变函数,并以颜色常数模型假设为基础,建立了描述彩色图像序列中运动模式的4元光流方程。为了对超复数4元光流方程中的运动参数进行有效估计,定义了4元同质微分的概念,进而得到了基于4元同质颜色常数模型的光流方程,并给出了该光流方程的优化求解方法。最后,与已有算法的对比实验验证了该文方法的有效性。     

基于四元数小波变换HMT模型的图像去噪

实小波变换域中的图像去噪会产生伪Gibbs现象,为此,提出一种基于四元数小波变换的隐马尔可夫树模型(Q-HMT),并应用于图像去噪。模型结合四元数小波变换理论与HMT模型,体现小波系数层间的相关性,捕获小波系数邻域的统计特征。实验结果表明,该模型的图像去噪效果在峰值信噪比以及视觉效果上均优于经典的去噪方法。     

基于多变换域的彩色图像多功能水印算法

针对单水印存在功能单一的问题,提出一种基于离散小波变换（DWT）和四元数离散余弦变换（QDCT）的彩色图像多功能水印算法。首先,将彩色图像分通道置乱后进行DWT,选择低、中频子带作分块QDCT,利用部分实数系数构造系数矩阵,鲁棒性水印通过加法原则嵌入该系数矩阵的奇异值中;然后,将图像进行2×2分块的QDCT,特征脆弱水印利用QDCT的低频模值系数产生,并嵌入空域最低有效位（LSB）。实验结果表明,鲁棒性水印具有良好的抗JPEG压缩、噪声、对比度调节、剪切、旋转以及混合攻击的能力,脆弱水印对篡改敏感且具有精确的篡改定位功能。     

分区四元数姿态控制

提出了一种基于分区控制策略的四元数姿态控制律. 其基本思想是基于姿态四元数误差分区设计目标角速度, 由此将问题降阶为一个角速度跟踪问题; 基于不同的角速度跟踪误差, 设计了切换类型的抗干扰姿态控制律. 该控制策略可以使得姿态快速收敛, 并且在合适的参数选择条件之下还能同时满足控制力矩的饱和约束. 通过综合相平面和Lyapunov函数的分析方法严格证明了闭环系统全局收敛的性质. 最后, 通过数值仿真验证了本文提出的控制方案的有效性.