共查询到19条相似文献，搜索用时 203 毫秒
1.
2.
基于边缘检测和投影法的车牌定位算法研究 总被引：3，自引：0，他引：3
车牌定位是车牌识别中的关键步骤。为了能在复杂背景和不同光照条件下快速、准确地定位车牌位置,提出了一种基于边缘检测和投影法的车牌定位方法。该方法首先对车牌图像实施边缘检测、二值化等预处理,然后在此基础上,利用基于双向回溯的投影法确定车牌的上下左右边界。实验结果表明,该方法定位速度快、准确率高。 相似文献
3.
4.
文章提出了一种基于灰度投影技术与霍夫变换的瞳孔定位方法:操作中首先对图像进行二值化处理,将特征点从人眼图像中分割出来,分别利用水平和垂直灰度投影,结合人眼的结构特征找到瞳孔的大致位置坐标,然后再利用基于圆的霍夫变换快速定位人眼.此方法大大提高了人眼定位的效率,实现了准确的瞳孔定位. 相似文献
5.
在电力高空作业应急救助中,救援绳索的质量是保证人员安全的重要因素之一,本文针对标准Hough变换算法在应用过程中还存在的检测精度不高的问题,提出了一种角度自适应积分投影的救援绳索缺陷检测机制。首先采用统计滤波对救援绳索进行视觉检测,提取救援绳索的灰度值图像,然后采用Hough变换算法对绳索的边界进行限定,并对其进行方向性膨胀和腐蚀的优化,接着采用角度自适应积分投影方法映射救援绳索内的连续性信息。算法仿真实验结果表明,本文提出改进算法相比较标准Hough变换算法,具有更高的检测精度,可以有效提高高空作业的安全性。 相似文献
6.
人脸检测作为人脸识别系统的重要一环,越来越受到技术研究和商业应用的关注。针对人脸检测中时间和检测率不能很好保证的情况,提出了使用DCT变换和支持向量机的人脸检测算法。利用离散余弦变换的系数作为支持向量机的输入特征值,证明该方法能提高人脸检测的准确性,并缩短检测时间。 相似文献
7.
提出一种人脸识别方法用于解决姿态变化对识别准确率的影响。首先检测人脸图像的SIFT特征,然后根据SIFT特征计算人脸图像间的多示例距离;基于此多示例距离,用保局投影将人脸图像映射至流形空间,最后在流形空间中采用K近邻方法进行人脸识别。该方法有三个特点：(1)采用SIFT特征减小了未知姿态对识别准确率的影响;(2)通过保局投影将特征变换到流形空间一个点,避免了复杂的SIFT特征匹配策略;(3)借助流形方法滤除高维特征中的噪声。实验结果表明与已有方法相比,在人脸姿态不确定的情况下,该方法能提供较为理想的识别准确率。 相似文献
8.
9.
疲劳驾驶和危险驾驶是造成交通事故的主要原因，文章基于YOLOv5目标检测算法和dlib人脸识别库，分别从人脸朝向、位置、眼睛开合度、眨眼频率、驾驶员手持物品定位，分析物品形态、物品大小等数据，通过这些数据，利用YOLOv5算法实时地计算出驾驶员是否存在疲劳驾驶和危险驾驶的行为，若存在则通过系统及时给出相应的安全提示。 相似文献
10.
11.
本文首先用投影法、截面法、柱面坐标法、球面坐标法[1]四种方法求解同一道三重积分计算实例,然后分析了各种方法最适合使用的条件和应注意的事项。 相似文献
12.
凑微分法和分部积分法是求解不定积分的两种重要方法,这两种方法都适用于形如乙∫f（x）g（x）dx的积分,都有凑微分的过程,学生容易混淆。针对这种情况,本文提出了一种简单、快速、易于学生掌握的方法——＂拆＂、＂判＂、＂选＂、＂凑＂四步法,帮助学生快速找到正确的积分方法和要凑的微分因子,从而达到快速、准确解题的目的。 相似文献
13.
研究文本定位与特征提取问题。针对传统的Canny算子图像检测算法的不足,提出了一种改进的Canny算子图像信息特征提取算法。研究方法是:首先对彩色图像进行高斯金字塔分解,然后用Canny算子检测彩色图像,提取边缘图像,再经过通二值化,去噪方差投影定位文本区域。实验结果表明,本文提出的方法有效、实用。 相似文献
14.
15.
针对灰度图像中的人脸检测问题,提出了一种基于多种支持向量机的决策融合检测方法。该方法首先用传统的二类支持向量机(C—SVM)和单类支持向量机(One-Class SVM)分别对图像进行检测,然后决策融合两种分类器的检测结果。在MIT CUM人脸库上的实验结果表明,该方法具有良好的检测效果和较低的虚警率。 相似文献
16.
Jay I. Frankel Majid Keyhani Rao V. Arimilli 《Journal of The Franklin Institute》2010,347(9):1681-1688
This paper derives a new integral relationship between heat flux and temperature in a transient, three-dimensional heat conducting Cartesian half space (x>0, y∈(−∞,∞), z∈(−∞,∞)). A unified mathematical treatment has been developed based on operational and transform methods; and singular integral equation regularization. Regularization is accomplished based on a series of observations involving the diffusive nature of the operator. This newly developed relationship provides the local heat flux perpendicular to the front surface at any location within the half space. This expression suggests that an embedded plane of temperature sensors parallel to the surface can be used to acquire the local, in-depth heat flux in the x-direction. The relationship does not require a priori knowledge of the surface boundary condition which has analytically been removed in the process. The ill-posed nature of diffusion is highlighted owing to the appearance of the heating/cooling rate (°C/s) in the integrand of the new relationship. Integral relationships of this type are highly useful for experimental investigations since the in-depth heat flux can be extracted from well-established temperature transducers. 相似文献
17.
18.
《Information processing & management》2023,60(3):103306
Most of the existing large-scale high-dimensional streaming anomaly detection methods suffer from extremely high time and space complexity. Moreover, these models are very sensitive to parameters,make their generalization ability very low, can also be merely applied to very few specific application scenarios. This paper proposes a three-layer structure high-dimensional streaming anomaly detection model, which is called the double locality sensitive hashing Bloom filter, namely dLSHBF. We first build the former two layers that is double locality sensitive hashing (dLSH), proving that the dLSH method reduces the hash coding length of the data, and it ensures that the projected data still has a favorable mapping distance-preserving property after projection. Second, we use a Bloom filter to build the third layer of dLSHBF model, which used to improve the efficiency of anomaly detection. Six large-scale high-dimensional data stream datasets in different IIoT anomaly detection domains were selected for comparison experiments. First, extensive experiments show that the distance-preserving performance of the former dLSH algorithm proposed in this paper is significantly better than the existing LSH algorithms. Second, we verify the dLSHBF model more efficient than the other existing advanced Bloom filter model (for example Robust Bloom Filter, Fly Bloom Filter, Sandwich Learned Bloom Filter, Adaptive Learned Bloom Filters). Compared with the state of the art, dLSHBF can perform with the detection rate (DR) and false alarm rate (FAR) of anomaly detection more than 97%, and less than 2.2% respectively. Its effectiveness and generalization ability outperform other existing streaming anomaly detection methods. 相似文献