结合特征置信度的背景感知相关滤波跟踪算法

基于相关滤波的跟踪算法表现突出,然而在复杂跟踪环境下跟踪器很容易发生预测错误而导致跟踪失败。针对这一问题,提出一种结合特征置信度的背景感知跟踪算法。首先,采用较大的样本采集区域捕捉背景信息对滤波器训练,利用二次项替换及交替方向乘子法对模型进行快速求解。在颜色特征、方向梯度直方图特征以及灰度特征等多特征融合的基础上,根据目标响应大小对每个特征通道动态计算特征置信度,通过特征置信度加权来训练新的滤波器。为提高滤波器鲁棒性,结合APCE和特征置信度对滤波器进行高置信度更新。所设计的算法在OTB100数据集上与多个优秀相关滤波算法进行对比,实验结果显示,该算法在多种复杂场景下均具有较好的准确性和实时性。

基于粗糙集的分类关联规则挖掘算法研究

本文给出了一种将属性约简和分类关联规则挖掘相结合的新型分类挖掘系统的算法(CARMA)。它运用粗糙集理论把关系数据库按属性值分成若干等价类、约简冗余属性及依赖属性,然后对数据约简后的目标关系表求取分类支持度大于阈值的强类和特征置信度大于阈值的强特征,从而有效获取强类中的强特征的决策关联规则。实验结果表明,CARMA对于数据的分类是有效的,比其它算法具有更高的分类精度和效率。它能够有效地克服ID3系列算法的冗余性、复杂性和对大数据量的不适应性,对增量数据能够达到较好的分类效果和具有广泛的应用前景。本文关键讨论了具体的算法、系统框架和实例。

智能车前方目标车辆检测方法研究

为了准确检测智能车前方目标车辆以及确定目标车辆与本车的距离信息,提出了一种应用模糊逻辑融合车辆多特征检测目标车辆的算法。该算法充分考虑到车辆多个特征的重要性,将模糊理论中的模糊隶属度函数和特征置信度值相结合实现目标车辆的验证。在生成目标假设以及提取有效特征点后,通过对不同特征进行模糊化,根据模糊推理融合得到的目标车辆的最终置信度验证车辆假设,实现前方车辆的检测。试验数据表明,该算法具有较好的适应能力和抗干扰能力,能够及时准确地检测前方目标车辆。

基于双重轮廓特征匹配的遮挡目标识别算法

针对传统目标识别算法对于遮挡目标匹配难度大、误匹配率高的问题,提出一种基于双重轮廓特征匹配的遮挡目标识别算法。首先,将形状轮廓分为采样点和轮廓有效片段两个层次,其中轮廓片段是利用离散曲线演化算法和分段优化算法获得的,用于表征形状的全局信息,而轮廓点的用于表征局部细节信息。其次,通过动态规整算法DTW(Dynamic Time Warping)衡量轮廓片段轮廓点之间的精确型傅里叶高度函数特征(Accurate Fourier Height Functions 2,AFHF2)的相似度,同时计算轮廓分段的置信度特征。最后,结合分段置信度再次匹配识别。实验结果表明,与常见的遮挡目标识别算法相比,所提出算法的检索精度有明显的提升。

Configurable ontology mapping based on multi-feature

A configurable ontology mapping approach based on different kinds of concept feature information is introduced in this paper. In this approach, ontology concept feature information is classified as five kinds, which respectively corresponds to five kinds of concept similarity computation methods. Many existing ontology mapping approaches have adopted the multi-feature reasoning, whereas not all feature information can be com- puted in the real ontology mapping and only fractional feature information needs to be selected in the mapping computation. Consequently a eonfigurable ontology mapping model is introduced, which is composed of CMT model, SMT model and related transformation model. Through the configurable model, users can conveniently select the most suitable features and configure the suitable weights. Simultaneously, a related 3-step ontology mapping approach is also introduced. Associated with the traditional name and instance learner-based ontology mapping approach, this approach is evaluated by an ontology mapping application example.

弱纹理环境下视觉里程计优化算法研究

针对视觉里程计在弱纹理的道路环境下定位精度骤降的问题,提出一种双目视觉里程计优化算法。首先,通过提取局部特征平面,为立体匹配缺失的特征提供平面约束,增加有效立体特征的数量;其次,在特征追踪过程中,使用匀加速运动模型,提高特征追踪的数量和质量;最后,在位姿计算和优化过程中,使用考虑特征置信度的光束法平差算法来减少远距离特征引入的误差影响,提高算法的精度和稳健性。数据集实验和实际场景实验表明,该算法在占有极少计算资源的情况下对弱纹理环境下的定位精度有较明显的优化效果,在其他场景也具有较好的适应性。

Effect of the uncertainty in soil-water characteristic curve on the estimated shear strength of unsaturated soil

Most failures or instabilities of geotechnical structures commonly result from shear failure in soil. In addition, many infrastructures are constructed within the unsaturated zone. Therefore, the determination of shear strength of unsaturated soil is crucial in geotechnical design. The soil-water characteristic curve(SWCC) is commonly used to estimate the shear strength of unsaturated soil because the direct measurement is time-consuming and costly. However, the uncertainty associated with the determined SWCC is rarely considered in the estimation of the shear strength. In this paper, the uncertainties of SWCC resulted from different factors are reviewed and discussed. The variability of the estimated shear strength for the unsaturated soil due to the uncertainty of SWCC associated with the best fit process is quantified by using the upper and lower bounds of the determined SWCC. On the other hand, the uncertainties of the estimated shear strength due to different initial void ratios or different confining pressures are quantified by adopting different SWCCs. As a result, it is recommended that the measured SWCC from the conventional Tempe cell or pressure plate needs to be corrected by considering different stress levels in the estimation of the shear strength of unsaturated soil.