提出了一种传统的指纹图像分割算法——方差法的改进算法。
An improved algorithm of the traditional fingerprint image segmentation algorithm-variance method is proposed.
针对汽车牌照与车身背景的分割问题，提出一种基于横向方差和数学形态学的图像分割方法。
To segment a vehicle license card from the image background of the whole vehicle, a image segmentation algorithm basing on cross variance and Mathematical morphology is presented.
对于海面图像，分别采用感兴趣舰船目标区域的方差值、目标和背景亮度对比度这两个特征对目标进行融合识别。
As to sea image, the variance feature of region of interest and the luminance contrast feature between target and background are used to fusion recognition.
在选择掩模平滑算法中，需要计算一幅图像的大量局部窗口的灰度均值和方差。
In selective masking smoothing algorithm, it is needed to calculate grayscale average and variance of a large number of image Windows.
首先，对机动车号牌图像进行大小、灰度方差、灰度均值的标准化处理。
First, we standardized the size, variance of gray scale and mean value of gray scale of the motor vehicle plate images.
在传统图像灰度方差评价函数的基础上，利用小波多分辨率分析，提出了一种新的自动聚焦算法。
On the basis of evaluation function of using image intensity variance, we proposed a new auto-focusing algorithm by means of Wavelet Transform Multi-Resolution Analysis(WTMRA) theory.
研究结果表明，烧伤表面图像的色度方差具有明显特征，可以作为烧伤程度的评价指标。
It is found that the variance in the chroma of the surface image is of significant characteristic, and that it can be used as an index to evaluate the burn degree.
在最大类间方差法和一致性准则法的基础上，运用最大熵原理来选择灰度阈值对图像进行分割。
Based on maximum between-cluster variance method and uniformity measure, this paper USES maximum entropy principle to select the gray-level threshold value for image segmentation.
最小类内方差法分割图像时需要计算二次统计量，运算量大，效率不高。
The method of minimum interclass variance needs to compute conic statistic in segmenting image, which costs a large computation but has less efficiency.
对子图像进行局部直方图计算，由最大方差比例阈值确定方法计算出子图像的最佳阈值。
Every child image of the local regions is calculated, and the best threshold value of child image with the method of the biggist proportion of squase deviations is calculated.
本文提出了一种基于二维最大类间方差的指纹图像分割算法。
A fingerprint image segmentation method based on 2d maximum between-cluster variance is suggested in this paper.
该方法利用动态规划理论中代价函数的建立过程来充分考虑图像中的灰度、均值、方差、光滑程度和纹理等多种特征，从而使目标检测的结果为最佳。
Using images' features such as gradient, mean, variance, degree of smoothing and texture characteristics to establish cost function may obtain the best results of target detection.
通过每一叶节点以及整个图像的噪声的估计方差变化计算叶节点阈值。
The threshold of each leaf is calculated through the variance of each and the estimated noise variance of the entire image.
给出了一种基于最大互信息和边缘互方差的医学图像配准算法。
Gives a new algorithm of medical image registration based on maximum of mutual information and edge correlative deviation.
根据模糊图像的频谱和协方差特性估计模糊图像的模糊度。
According to the frequency spectrum and covariance, the blur strength of the blurred image is evaluated.
利用最大类间方差法（OTSU）求出对图像进行二值化处理的最佳阈值，从而进行图像二值化处理。
Using the maximum between-class variance method (OTSU) obtained image binarization best treatment threshold, and thus image binarization treatment.
在计算信号的协方差矩阵时，对图像分块并用椭圆窗滑动求局部协方差矩阵，以达到局部自适应的去噪目的。
When the signal covariance is calculated, the image is cut into several blocks and the elliptic window is used, so the local covariance is achieved.
针对图像中的高灰度区和灰度剧变区，应用图像局部均值和方差自适应调节增强系数。
For the regions in image where gray value is high or change intensely, local mean value and variance is adopted to control enhancement coefficients.
最大类间方差法的二值化算法是基于类间方差达到最大为准则，采用该法获得了理想的二值化效果，较准确地从岩心剖面中提取出裂缝图像信息。
The optimal effect of binarization arithmetics is the maximum variance between various classifications. Core fissure image is acquired from core splitting surface images by the use of this method.
我们在分析参数活动轮廓和几何活动轮廓模型的基础上，提出最小方差参数活动轮廓模型，并成功应用于医学图像自动分割。
After analyzing snake and geometric active contour model, a minimum variation snake model was proposed and successfully applied to weak edge medical image segmentation.
提出了一种新的基于一维最大类间方差和形状连通性的图像分割算法。
A novel image segmentation algorithm based on 1-d maximum inter-class variance and shape connectivity is proposed in this paper.
针对图像中的高灰度区和灰度剧变区，应用图像局部均值和方差自适应调节增强系数。
For regions with high grey-scale level or with tensely variable grey level, the image enhancement coefficients shall be adaptively controlled based on local mean value and variance.
提出了利用边缘点集的协方差矩阵的特征值与特征矢量作为人脸图像尺度与方向的粗估计方法，从理论和实验上证明了该方法的可行性。
A new method for human face image scale and direction estimation is proposed using covariance matrix eigenvalue and eigenvector of edge set, which is proved by theory and experiments.
该算法以低通滤波算法估计背景中的低频成分，及图像局部方差估计背景起伏，来计算出局部自适应的目标检测判决门限。
The algorithm adopts low pass filter to estimate the background and calculates the standard deviation to estimate the fluctuation of background to get the local adaptive detection threshold.
个体指纹被采集成指纹图像后，首先采用基于自身灰度均值和方差的规格化方法，降低指纹脊、谷线之间的灰度变化程度；
Based on the mean value and variance, the image normalizing is to reduce the variations of gray between fingerprint ridges and furrows;
给出了该滤波方法的噪声方差方程，并利用该方法对采集的视频图像数据进行了处理，讨论了不同阈值下的滤波效果。
The noise variance equation is demonstrated here and is used to sort out the accumulated data of the video images. The filtering effects under different thresholds are also discussed here.
接着根据靶标图像和光斑图像在灰度特征上的差异，采用了最大类间方差自适应分割法实现了光靶图像的分割。
Thirdly, according to the differences of target images and spot images in gray-scale features, Otsu adaptive segmentation method is adopt to realize the image segmentation.
针对传统的基于方差法的指纹图像分割法存在的问题，提出了一种改进的指纹图像分割算法。
Address the traditional variance-based method for fingerprint image segmentation problems, put forward an improved algorithm for fingerprint image segmentation.
用统计学的方法，等角度间隔地计算归一化步态轮廓图像各像素点至质心距离的均值与方差，并用其构造步态识别的特征向量。
Using statistical methods to calculate the Angle interval mean and variance of pixels to centroid distance of the normalized gait silhouette images, and construct them as a feature vector.
用统计学的方法，等角度间隔地计算归一化步态轮廓图像各像素点至质心距离的均值与方差，并用其构造步态识别的特征向量。
Using statistical methods to calculate the Angle interval mean and variance of pixels to centroid distance of the normalized gait silhouette images, and construct them as a feature vector.
应用推荐