Description: 边缘特征的提取就是求图像梯度的局部最大值和方向。实际计算中,以微分算子的形式表示,并采用快速卷积函数来实现。常用的算子有微分算子,拉普拉斯算子,Canny算子等。其中Canny边缘检测是一种较新的边缘检测算子,具有较好的边缘检测性能,得到越来越广泛的应用。Canny边缘检测法利用高斯函数的一阶微分,它能在噪声抑制和边缘检测之间取得较好的平衡-Edge feature extraction is to seek the local maximum of image gradient and orientation. The actual calculation to the form of differential operator representation, and using fast convolution function to achieve. Commonly used operators are differential operators, Laplace operator, Canny operator and so on. Canny edge detection which is a relatively new edge detection operator, and has good edge detection performance, get more and more widely used. Canny edge detection method using first derivative of Gaussian function, it can in the noise suppression and edge detection to achieve a better balance between Platform: |
Size: 1024 |
Author:xiaowei |
Hits:
Description: Example of a two-dimensional Gabor filterIn image processing, a Gabor filter, named after Dennis Gabor, is a linear filter used for edge detection. Frequency and orientation representations of Gabor filter are similar to those of human visual system, and it has been found to be particularly appropriate for texture representation and discrimination. In the spatial domain, a 2D Gabor filter is a Gaussian kernel function modulated by a sinusoidal plane wave. The Gabor filters are self-similar – all filters can be generated from one mother wavelet by dilation and rotation
Platform: |
Size: 7168 |
Author:Real1 |
Hits:
Description: Contourlet变换是一种新的图像二维表示方法,具有多分辨率、局部定位、多方向性、近邻界采样和各向异性等性质,其基函数分布于多尺度、多方向上,少量系数即可有效地捕捉图像中的边缘轮廓,而边缘轮廓正是自然图像中的主要特征。-Contourlet image transformation is a new two-dimensional representation, with multi-resolution, local orientation, multi-directional, neighboring community and anisotropy nature of sampling, the basis function distribution in the multi-scale, multi-up, a few factors that can effectively capture the image edges, and edges is the main features of natural images. Platform: |
Size: 1343488 |
Author:梅雪 |
Hits:
Description: Utilisation des certaines critères comme les textures ,la couleur et la forme ,donc en peut associer un vecteurs de caractéristiques pour l’image qui se trouve dans notre base des données , Nous utilisons les maxima de courbure de passage a zéros contours de Curvature Scale Space (CSS) sur l’image comme un vecteur de caractéristiques , l’approche consiste a calculer l’image CSS de chaque frontière et ensuite trouver les maxima des contours CSS comme un discripteur de forme pour comparer des images , les propriétés de base da la representation CSS qui rendent presque unique pour la recherche de similarité sont les suivant :
- Il est robuste par rapport au bruit, l’échelle et l’orientation de l’objet
- Il conserve les informations locales de la forme d’entrée. Chaque concavité ou convexité de la forme a ses propres correspondants contour sur l’image CSS, chaque point dans l’axe horizontal de l’image de CSS à son correspondant sur les limites réelles.-Utilisation des certaines critères comme les textures ,la couleur et la forme ,donc en peut associer un vecteurs de caractéristiques pour l’image qui se trouve dans notre base des données , Nous utilisons les maxima de courbure de passage a zéros contours de Curvature Scale Space (CSS) sur l’image comme un vecteur de caractéristiques , l’approche consiste a calculer l’image CSS de chaque frontière et ensuite trouver les maxima des contours CSS comme un discripteur de forme pour comparer des images , les propriétés de base da la representation CSS qui rendent presque unique pour la recherche de similarité sont les suivant :
- Il est robuste par rapport au bruit, l’échelle et l’orientation de l’objet
- Il conserve les informations locales de la forme d’entrée. Chaque concavité ou convexité de la forme a ses propres correspondants contour sur l’image CSS, chaque point dans l’axe horizontal de l’image de CSS à son correspondant sur les limites réelles. Platform: |
Size: 973824 |
Author:oussama |
Hits:
Description: 基于纹理和边缘方向特征融合的目标跟踪程序代码,是基于mean shift框架的-This paper proposes a powerful and robust local descriptor,called textureorientation
descriptor(TOD). TOD consists of two components: texture and
orientation. Considering that the human shape shows more short, fragmentary
and vertical edges, the proposed TOD descriptor encodes the spatial information
of pedestrian image. We combine the proposed TOD representation
model with mean shift framework to perform pedestrian tracking. Results
from experiments show the improved performance of the proposed algorithm,
in the scenarios of heavy occlusions, illumination variations and similar
objects interference, when compared with the other descriptors based trackers. Platform: |
Size: 5120 |
Author:邹温林 |
Hits:
Description: Thoughtworks公司面试题——MARS ROVERS问题火星探测器 C# 实现 VS2010工程,带界面展示!
一小队机器人探测器将由NASA送上火星高原,探测器将在这个奇特的矩形高原上行驶。
用它们携带的照相机将周围的全景地势图发回到地球。每个探测器的方向和位置将由一个x,y系坐标图和一个表示地理方向的字母表示出来。为了方便导航,平原将被划分为网格状。位置坐标示例:0,0,N,表示探测器在坐标图的左下角,且面朝北方。为控制探测器,NASA会传送一串简单的字母。可能传送的字母为: L , R 和 M 。 L ,和 R 分别表示使探测器向左、向右旋转90度,但不离开他所在地点。 M 表示向前开进一个网格的距离,且保持方向不变。假设以广场(高原)的直北方向为y轴的指向。
输入:首先输入的line是坐标图的右上方,假定左下方顶点的坐标为0,0。剩下的要输入的是被分布好的探测器的信息。每个探测器需要输入wo lines。第一条line 提供探测器的位置,第二条是关于这个探测器怎样进行高原探测的一系列说明。位置是由两个整数和一个区分方向的字母组成,对应了探测器的(x,y)坐标和方向。每个探测器的移动将按序完成,即后一个探测器不能在前一个探测器完成移动之前开始移动。-he Thoughtworks Company interview questions- MARS ROVERS Mars probe C# VS2010 project with interface shows!
The robot detector by a small team of NASA Mars plateau, the detector will be traveling on this strange rectangular plateau.
They carry the camera panoramic views of the surrounding terrain Figure is sent back to Earth. The direction and position of each of the detectors by an x, y-based coordinate diagram, and a representation of the geographical direction of letters represented. In order to facilitate navigation, the plains will be divided into a grid-like. Sample location coordinates: 0,0 N detectors in the lower left corner of the plot, and facing north. Control detectors, NASA will send a bunch of letters. The letters may be sent to: L , R and M . L , R , respectively, so that the detector rotated 90 degrees to the left, right, but does not leave his location. M is moved into a grid distance forward and changing the orientation. Assuming Square (plateau) straig Platform: |
Size: 54272 |
Author:wanghu |
Hits:
Description: 本文提出一种通过实时调整目标特征权值来进行背景自适应跟踪的算法。首先,定义了一种综合特征集合用以描述目标的颜色和局部轮廓。其次,提出了在滤波框架中对目标特征进行评估的算法,从而使得具有强区分能力的特征占有较大的权值,进而使其能够在跟踪过程起到较大的作用。采用传统的Kalman 滤波和粒子滤波对所提出的算法进行了验证。-In this paper, we propose a new adaptive visual object tracking method based on
online feature evaluation approach. First, a feature set is built by combining color
histogram (HC) with gradient orientation histogram (HOG), which emphasizes both
color and contour representation. Then a feature confidence evaluation approach is
proposed to make features with higher confidences play more important roles in the
instantaneous tracking ensuring that the tracking can adapt to the appearance change
of both the object and its background. The feature evaluation approach is fused with
filter frameworks, e.g. Kalman and Particle filter, to keep the temporal consistency of feature confidence evolution. Platform: |
Size: 1602560 |
Author:wenping |
Hits: