基于L~1范数变分模型的高密度椒盐噪声滤波

针对现有图像滤波算法在去除高密度椒盐噪声方面的不足,提出了一种基于L1范数变分模型的去噪算法.该算法首先根据椒盐噪声的特点和像素的局部灰度特征分离出噪声点和信号点,在滤波过程中,对信号点不予处理,而对噪声点采用基于L1范数的变分模型进行恢复.由于椒盐噪声的灰度值与原始像素无关,因此,去噪时只利用噪声点邻域内信号点本身的灰度信息,并将已处理过的噪声点当作新的信号点,以避免对下一像素滤波时的影响.最后通过仿真实验,结果表明,在高密度噪声条件下(50%),该算法的噪声去除能力和细节保护能力均可获得令人满意的结果.     

基于嵌入式的食品中重金属光电检测系统设计

基于光电传感方法和MSP430单片机控制技术实现对食品中的重金属进行快速检测。介绍了光电传感检测方法的基本原理和系统光控制电路设计,使光强按照不同波长自动调整到最佳状态并稳定不变,解决了传统分光检测系统在恶劣环境条件下性能下降的问题。实验对比分析说明,该检测系统具有良好的准确性和可靠性,可对现场食品样品进行快速粗筛,节省了测试时间。     

嵌入式实时车道线偏移快速侦测算法及系统实现

以CortexA9处理器OMAP4460作为核心处理器,并在此基础上提出一种算法,设计了一种嵌入式实时侦测车道线偏移的系统。实验表明,该系统在路况复杂及不同光强干扰的情况下仍能表现出较强的鲁棒性。     

泵送粉煤灰混凝土回弹测强曲线特性

通过对不同强度等级、不同粉煤灰掺量的泵送混凝土标准立方体试块进行抗压、回弹、碳化试验,研究各测试数据随龄期的变化规律,根据常用曲线形式进行测强曲线拟合并分析其与普通混凝土统一测强曲线的区别．试验表明：混凝土各测试数据的发展受粉煤灰影响不同而呈现不同规律,掺量高的低强混凝土抗压强度发展慢,回弹发展快,碳化深度大;而掺量低的高强混凝土抗压强度发展快,回弹发展慢,碳化深度低．该现象的内在原因为掺量低时粉煤灰微集料效应为主,抗压强度提高明显;掺量高时粉煤灰低活性限制抗压强度增长,而回弹值提高明显．各测试数据的发展规律不同造成曲线拟合误差大,分析建议将胶凝材料含量〉400kg／m2、粉煤灰掺量≤20％的高强混凝土的回弹值适当增大后拟合,可得到满足地区曲线的精度要求的统一曲线．     

基于改进CGHT的光学遥感图像舰船尾迹检测

根据遥感图像舰船目标检测的实时性要求,提出了一种应用改进CGHT算法检测舰船尾迹的方法.通过加入舰船船体的检测结果,改进CGHT算法中种子点的选取方式,以改善尾迹检测的计算复杂度,并将检测重点侧重于舰船船体附近区域,避免尾迹检测的盲目性.实验表明,该算法显著减少了尾迹检测的运行时间,降低了伪尾迹引起的虚警,提高了尾迹定位的准确性.     

物联网中重复博弈论入侵检测模型

由于物联网通信节点自身的缺陷,导致现有的安全方法难于应用到物联网中。针对这一情况,提出基于重复博弈论的入侵检测模型。建立一种用于检测恶意通信节点的重复博弈模型算法,使模型应用更符合实际情况,结合随机最优反应均衡算法优化模型使结果能更适应实际网络并且能趋于纳什均衡,并引入一种通用惩罚策略,刺激通信节点之间采取合作策略,提高模型发包成功率。通过对模型算法进行实验仿真,表明该模型能够有效地遏制恶意节点的攻击,提高网络效率。     

基于动态区域规划的双模型车道线识别方法

为提高车道偏离预警以及前方车辆识别系统性能,提出一种基于动态区域规划的双模型结构化道路车道线检测方法。通过基于模板的腐蚀、膨胀处理改善了图像预处理效果。根据“微元”理论进行图像动态细化分区,对各区域局部独立识别,整体综合分析,最终利用直线和B-样条曲线双模型完成车道线拟合。 实车试验证明,直线/B-样条曲线双模型拟合,提高了算法对不同车道线的适应性。      

多通道脉冲激光探测技术研究

在介绍多通道激光探测系统工作原理的基础上,设计了一种基于数字化的激光探测系统,采用一种基于抛物线内插互相关时延估计的激光定距方法,从而进一步提高定距精度. 该系统采用6路双通道ADC并行采样,具有较好的实时性. 仿真结果表明,互相关时延估计算法能够在低信噪比（R_SN=-7 dB）下实现定距,最大距离误差为0.74 m.      

A machine learning-based method to detect fluorescent spots and an accelerated,parallel implementation of this method

Under most microscopes, the fluorescent sig- nals emitted from biological structures, such as vesicles, as well as the signals from single molecules will appear as small puncta, which contribute to a Gaussian-like distri- bution. Accurate segmentation of these spots will funda- mentally affect our interpretation of a specific biological progress. Because of the complicated backgrounds in images, many algorithms fail to identify all of the inter- esting signals; the tremendous amount of time required for algorithms to process large datasets can also decrease their utility. Here, we introduce an excellent robust detection method based on the machine learning algorithm Ada- Boost, which outperforms threshold-based segmentation, wavelets, and FDA under most situations. We also provide a GPU/multi-core CPU implementation of this algorithm; this implementation accelerates the algorithm approxi- mately 10- and 7-fold acceleration compared with a single CPU implementation. The great reduction of time should make this method a promising candidate in the processing of large datasets. Furthermore, we demonstrate the use of our algorithm on true fluorescent micrographs, and the results show that machine learning-based detection meth- ods outperform the four other previously reported methods.     

On design of physical random access channel in high-speed railway LTE systems

For the uplink in the long-term evolution （LTE） systems, the physical random access channel （PRACH） adopts Zadoff-Chu （ZC） sequences as preamble sequences, and each user can choose a unique cyclic shift for user identification. In high-mobility scenario, the Doppler fre- quency shift raises up the false alarm probability by inducing the correlation peak offsetting at the receiver side. Traditional detection method utilizes multiple search win- dows for searching the potential uplink users. When the Doppler frequency shift is larger than the subcarrier spac- ing, the energy of correlation peak leaks outside the search window and increases the false alarm probability. In this paper, we present a novel design for PRACH preamble sequence for high-speed railway scenario. Through numerical observation, we find out that the offset of the correlation peaks only depends on the initial root sequence number and the length of ZC sequences. Therefore, the root number of ZC sequences should be carefully designed to generate a preamble sequence to fight against the Doppler frequency shift. Consequently, those offsets of the corre- lation peaks are concentrated within a single search win- dow, resulting in lower complexity of detection and lower probability of false alarm for random access in LTE uplink.