共查询到18条相似文献,搜索用时 140 毫秒
1.
2.
3.
4.
5.
6.
为了克服椭圆检测过程中对椭圆完整性和边缘梯度信息依赖性较强的缺点,实现椭圆目标的快速有效检测,提出了一种新的基于弧段提取的椭圆拟合方法。首先将梯度方向符号相同的相邻边缘点连接成弧段,然后根据弧段的凸性和象限分类组成包含三段弧的弧组,定义了新的弧选择策略,利用位置约束和弧对的椭圆中心估计提取候选椭圆,最后采用改进的拟合算法拟合椭圆。实验结果表明,基于弧段提取的椭圆拟合方法相对于LMEDS算法和RHT3具有更好的准确性、鲁棒性和稳定性,实时性也有一定的提高。 相似文献
7.
8.
为了避免基于有限更新率的连续域信号采样涉及复杂的运算问题,认真研究分段线性序列特点,提出一种在离散域进行信号参量提取的方法,称之为序列重采样.给出了有限长序列更新率定义,总结出一种适合简单分段线性序列的参量提取算法——试探法,将之推广到一般分段线性序列.不需要事先知道更新率的数量,也不需要在连续域进行运算,从采样后的序列中直接检测信号更新点和更新参量,从而实现信号采样.对所提出的算法进行了仿真,结果表明该算法是有效的. 相似文献
9.
10.
对转炉炼钢终点的实时精准控制能够有效提高钢铁产出的质量,炉口火焰光谱在炼钢不同时期的变化明显,对其进行分析处理并与机器学习方法相结合可有效用于炼钢终点的实时控制。针对炉口火焰光谱数据量大、现有方法对光谱特征提取在可信度和实时性上不足的缺陷,提出一种基于窗口竞争性自适应重加权采样(WCARS)结合迭代式连续投影算法(ISPA)的光谱特征波长选择方法,该方法在有效解决模型过拟合问题的同时,能够降低高维数据计算的复杂度。将火焰光谱数据沿波长方向进行窗口划分后,使用CARS进行计算选出特征窗口波段,再将迭代式选择与传统连续投影算法相结合,通过重复迭代精选出特征波长,在此基础上使用支持向量机回归(SVR)建立炼钢终点碳含量预测模型。实验采集363组炼钢后期的炉口火焰光谱数据作为样本,并对其进行Savitzky-Golay平滑预处理。使用WCARS-ISPA算法从全光谱数据中选出10个特征波长作为SVR模型的输入,碳含量为模型输出,Kennard-stone算法对训练集和测试集进行划分,选择碳含量的平均预测误差、预测误差在±2%以内的命中率以及运行30次的平均时间作为模型评价指标。实验结果显示,模型的平均碳含量预测误差为1.413 2%,命中率高达90.63%,运行时间为0.019 679 s。与使用全光谱和WCARS-ISPA,CARS-SPA,WCARS和SPA四种不同特征选择方法选出的特征波长建模得到的结果进行对比,基于WCARS-ISPA方法选出的特征波长建立的终点碳含量预测模型误差最小、命中率最高。提出一种新的炉口火焰光谱特征波长提取方法,使用窗口竞争性自适应重加权采样结合迭代式连续投影算法选取特征波长,并在此基础上建立转炉炼钢终点碳含量预测模型,实验结果表明,该方法能够有效提取火焰光谱特征,所建模型能够对转炉炼钢终点进行准确预测,满足工业生产的实时控制要求,为实际生产提供可靠帮助。 相似文献
11.
光学测量中椭圆圆心定位算法研究 总被引:9,自引:1,他引:8
提出了一种椭圆圆心定位算法。先对图像中的椭圆目标进行粗定位,获取椭圆的粗定位信息,再用Canny算子对图像中的感兴趣区域进行边缘检测。根据椭圆的粗定位信息以及图像的边缘灰度分布特征,用高斯曲线拟合的方法求得精确的亚像素边缘点的坐标。针对亚像素边缘点中出现的“孤立点”以及噪声点,分别用曲率滤波和均值滤波的方法加以滤波,将滤波后的亚像素边缘点用最小二乘法加以拟合,求得精确的椭圆圆心数据。实验证明了该算法的定位精度和抑制噪声能力,通过测试算法运行时间,证明该算法具有很好的实时性。 相似文献
12.
Freund I 《Optics letters》2004,29(15):1715-1717
The canonical point singularity of elliptically polarized light is a C point, an isolated point of circular polarization surrounded by a field of polarization ellipses. The defining singular property of a C point is that the surrounding ellipses rotate about the point. It is shown that this rotation is seen only for a particular line of sight (LOS) and, conversely, that there exists a unique LOS for every ellipse along which the ellipse is seen as a singularity. It is also shown that changes in LOS can turn singularities into stationary points and vice versa. The democratic behavior of polarization singularities and stationary points is a consequence of the fundamental "what you see is what you get" property of ellipse fields. Simple experiments are proposed for observing this unusual property of elliptically polarized light. 相似文献
13.
L. Li 《Opto-Electronics Review》2012,20(2):120-125
In this paper, an accurate 3D model analysis of a circular feature is built with error compensation for robot vision. We propose
an efficient method of fitting ellipses to data points by minimizing the algebraic distance subject to the constraint that
a conic should be an ellipse and solving the ellipse parameters through a direct ellipse fitting method by analysing the 3D
geometrical representation in a perspective projection scheme, the 3D position of a circular feature with known radius can
be obtained. A set of identical circles, machined on a calibration board whose centres were known, was calibrated with a camera
and did the model analysis that our method developed. Experimental results show that our method is more accurate than other
methods. 相似文献
14.
A novel two-dimensional (2D) pattern used in camera calibration is presented. With one feature circle located at the center, an array of circles is photo-etched on this pattern. An ellipse recognition algorithm is proposed to implement the acquisition of interest calibration points without human intervention. According to the circle arrangement of the pattern, the relation between three-dimensional (3D) and 2D coordinates of these points can be established automatically and accurately. These calibration points are computed for intrinsic parameters calibration of charge-coupled device (CCD) camera with Tsai method. A series of experiments have shown that the algorithm is robust and reliable with the calibration error less than 0.4 pixel. This new calibration pattern and ellipse recognition algorithm can be widely used in computer vision. 相似文献
15.
Yevhen Havrylenko Yuliia Kholodniak Serhii Halko Oleksandr Vershkov Oleksandr Miroshnyk Olena Suprun Olena Dereza Taras Shchur Mcisaw
rutek 《Entropy (Basel, Switzerland)》2021,23(7)
The problem of modelling a smooth contour with a regular change in curvature representing a monotone curve with specified accuracy is solved in this article. The contour was formed within the area of the possible location of a convex curve, which can interpolate a point series. The assumption that if a sequence of points can be interpolated by a monotone curve, then the reference curve on which these points have been assigned is monotone, provides the opportunity to implement the proposed approach to estimate the interpolation error of a point series of arbitrary configuration. The proposed methods for forming a convex regular contour by arcs of ellipses and B-spline ensure the interpolation of any point series in parts that can be interpolated by a monotone curve. At the same time, the deflection of the contour from the boundaries of the area of the possible location of the monotone curve can be controlled. The possibilities of the developed methods are tested while solving problems of the interpolation of a point series belonging to monotone curves. The problems are solved in the CAD system of SolidWorks with the use of software application created based on the methods developed in the research work. 相似文献
16.
近红外光谱(NIRS)在定量和判别分析中已得到广泛应用,化学计量学在其中发挥了重要作用,但仍需要建立基于新原理的方法,简化数据处理和建模过程,使近红外光谱分析更加方便、更加快速。多元光学计算(MOC)技术通过设计合适的光学滤波器可以在光谱测量的同时,根据光谱的整体形状得到定性定量结果。作为一种新的测量和计算方式,近年来在光谱分析领域逐渐得到应用。基于多元光学计算的原理,基于主成分分析和Fisher判别准则设计了近红外光谱的判别滤波器,将近红外光谱投影到二维空间,并在二维空间中计算每一类样品的置信椭圆作为模型进行判别分析。预测样本在二维空间的投影与模型的距离可以作为判别参数,判别值小于等于1时预测样品与模型样品判别为同一类别,否则判别为不同类别,且距离越大,差异性越大。采用460个不同部位的烟叶样品和73个不同生产厂家的药品对所建立的方法进行了测试,表明了方法的准确性。对于三类不同部位烟叶样品和四类不同厂家生产的药品,预测结果的真阳性率可以达到90%以上(除上部烟叶样品外),药品的真阳性率高达95%以上。但烟叶样品的假阳性率仍有些偏高,对于光谱极为相似的实际生产样品结果仍属可接受范围。所建立的方法可推广到其他应用领域,广泛用于基于近红外光谱的质量控制、产品检测、生产一致性监控等。 相似文献
17.
18.
用黑斑的椭圆性质在二值化图像中探测眼睛 总被引:3,自引:0,他引:3
在人脸探测过程中 ,双眼的探测非常重要 ,这是因为人的眼睛是人脸上最容易探测到的特征部位。通过双眼探测还可以对人脸进行定位 ,使人脸的尺寸归一化。探讨了一种基于二值化图像中黑斑几何形状分析的双眼探测方法。因为人眼的形状比较接近椭圆 ,故在研究黑斑的几何性质时 ,重点研究黑斑的椭圆性质。所谓黑斑的椭圆性质是指按照黑斑的面积和二阶矩计算出的椭圆参数。椭圆的主要参数有椭圆的面积、质心、长短轴的长度和方向角 (椭圆长轴与水平方向的夹角 )。为了描述黑斑与椭圆的近似程度 ,定义了椭圆硬度的概念。根据眼睛黑斑椭圆参数的测试数据建立了探测眼睛的数学模型 ,并给出了根据此模型得到的双眼探测的实验结果 相似文献