搜狗做题网翻译一段有关计算机方面的英文
来源：学生作业帮 编辑：搜狗做题网作业帮 分类：英语作业 时间：2024/06/04 09:34:40
翻译一段有关计算机方面的英文
Background representation
The MoG representation can be in RGB space,but also other color spaces can be applied,see for an overview.Often a representation where the color and intensities are separated is applied,e.g.,YUV,HSV,and normalized RGB,since this allows for detecting shadow-pixels wrongly classified as object-pixels.Using a MoG in a 3D color space corresponds to ellipsoids or spheres(depending on the assumptions on the covariance matrix) of the Gaussian representations.Other geometric representations are truncated cylinders and truncated cones.
Conceptually different representations have also been developed.Elgammal et al.use a kernel-based approach where they represent a background pixel by the individual pixels of the last N frames.Haritaoglu et al.represent the minimum and maximum value together with the maximum allowed change of the value in two consecutive frames.Heikkila and Pietikainen represent each background pixel by a bit sequence.where each bit reflects whether the value of a neighboring pixel is above or below the pixel of interest,i.e.,a texture operator.This makes the background model invariant to monotonic illumination changes.Oliver et al.also use a pixel's neighbors to represent it.They apply an eigenspace representstion of the backgrounf and detect new objects by comparing the input image with an image reconstructed via the eigenspace.
Eng et al.divide a background model learnt over time into a number of non-overlapping blocks.The pixels within each block are grouped into at most three classed according to homogeneity.The means of these classes are then the representation of the background for this block,i.e..a spatio-temporal block-based(overlapping blocks)background segmentation.Other spatio-temporal approaches are and where the background is represented by a predicted region found by an autoregresive process.
要详细的翻译.非常感谢^_^
Background representation
The MoG representation can be in RGB space,but also other color spaces can be applied,see for an overview.Often a representation where the color and intensities are separated is applied,e.g.,YUV,HSV,and normalized RGB,since this allows for detecting shadow-pixels wrongly classified as object-pixels.Using a MoG in a 3D color space corresponds to ellipsoids or spheres(depending on the assumptions on the covariance matrix) of the Gaussian representations.Other geometric representations are truncated cylinders and truncated cones.
Conceptually different representations have also been developed.Elgammal et al.use a kernel-based approach where they represent a background pixel by the individual pixels of the last N frames.Haritaoglu et al.represent the minimum and maximum value together with the maximum allowed change of the value in two consecutive frames.Heikkila and Pietikainen represent each background pixel by a bit sequence.where each bit reflects whether the value of a neighboring pixel is above or below the pixel of interest,i.e.,a texture operator.This makes the background model invariant to monotonic illumination changes.Oliver et al.also use a pixel's neighbors to represent it.They apply an eigenspace representstion of the backgrounf and detect new objects by comparing the input image with an image reconstructed via the eigenspace.
Eng et al.divide a background model learnt over time into a number of non-overlapping blocks.The pixels within each block are grouped into at most three classed according to homogeneity.The means of these classes are then the representation of the background for this block,i.e..a spatio-temporal block-based(overlapping blocks)background segmentation.Other spatio-temporal approaches are and where the background is represented by a predicted region found by an autoregresive process.
要详细的翻译.非常感谢^_^
背景表示法
MoG 表示法可能是在RGB 空间,而且其它彩色空间可能是应用的,看为overview.Often 颜色和强度被分离是应用的,即,YUV 、HSV,和正常化的RGB 的表示法,因为这考虑到查出阴影映像点错误地被分类当对象pixels.Using MoG 在3D 彩色空间对应于椭球或spheres(depending 在假定在协变性矩阵) 高斯representations.Other 几何学表示法是被削的圆筒和被削的锥体.
概念性地不同的表示法并且是developed.Elgammal 和一种基于仁的方法他们代表背景映像点由最后N 各自的映像点frames.Haritaoglu 和al.represent 极小和最大值与最大值一起允许价值的变动在二连贯frames.Heikkila 和Pietikainen 代表各个背景映像点由位sequence.where 各位的al.use 反射在是否之上一个邻居映像点的价值是或在映像点利益,即,纹理之下operator.This 使背景模型不变式对单调照明changes.Oliver 和al.also 使用pixel's 邻居代表it.They 申请backgrounf 的一eigenspace representstion 并且查出新对象由输入图象与图象比较被重建通过
背景模型学会随时间入一定数量的非重复blocks.The 映像点在各个块之内被编组入至多三分类根据这些类homogeneity.The 手段的英语和al.divide 是然后背景的表示法为这个块,i.e..a spatio-temporal 阻拦based(overlapping blocks)background segmentation.Other spatio-temporal 方法是和背景由一个被预言的区域代表由一个autoregresive 过程发现的地方.
MoG 表示法可能是在RGB 空间,而且其它彩色空间可能是应用的,看为overview.Often 颜色和强度被分离是应用的,即,YUV 、HSV,和正常化的RGB 的表示法,因为这考虑到查出阴影映像点错误地被分类当对象pixels.Using MoG 在3D 彩色空间对应于椭球或spheres(depending 在假定在协变性矩阵) 高斯representations.Other 几何学表示法是被削的圆筒和被削的锥体.
概念性地不同的表示法并且是developed.Elgammal 和一种基于仁的方法他们代表背景映像点由最后N 各自的映像点frames.Haritaoglu 和al.represent 极小和最大值与最大值一起允许价值的变动在二连贯frames.Heikkila 和Pietikainen 代表各个背景映像点由位sequence.where 各位的al.use 反射在是否之上一个邻居映像点的价值是或在映像点利益,即,纹理之下operator.This 使背景模型不变式对单调照明changes.Oliver 和al.also 使用pixel's 邻居代表it.They 申请backgrounf 的一eigenspace representstion 并且查出新对象由输入图象与图象比较被重建通过
背景模型学会随时间入一定数量的非重复blocks.The 映像点在各个块之内被编组入至多三分类根据这些类homogeneity.The 手段的英语和al.divide 是然后背景的表示法为这个块,i.e..a spatio-temporal 阻拦based(overlapping blocks)background segmentation.Other spatio-temporal 方法是和背景由一个被预言的区域代表由一个autoregresive 过程发现的地方.