多尺度分割原理与应用
合集下载
相关主题
- 1、下载文档前请自行甄别文档内容的完整性，平台不提供额外的编辑、内容补充、找答案等附加服务。
- 2、"仅部分预览"的文档,不可在线预览部分如存在完整性等问题,可反馈申请退款(可完整预览的文档不适用该条件!)。
- 3、如文档侵犯您的权益，请联系客服反馈,我们会尽快为您处理(人工客服工作时间：9:00-18:30)。
available providing a large amount of images for PSv: arious thematic applications.
维数灾难(Curse of Dimensionality):通常是指在涉
Introduction
The Curse of Dimensionality of hyperspectral remote sensor technology :
Advantages of hyperspectral remote sensor technology:
➢The detailed spectral information increases the possibility of more accurately discriminating materials of interest.
➢the rate of convergence of the statistical estimation decreases when the dimension grows while conjointly the number of parameters to estimate increases, making the estimation of the model parameters very difficult.
Hierarchical SegmHeSnegtation
Reading Report
CONTENTS
PART ONE
PARTБайду номын сангаасTWO
PART THREE
About the Introduction. The Hseg Segmentation. Application in ENVI.
1 PART OAbNouEt the Introduction.
➢with a limited training set, beyond a certain limit,
Introduction
How to build accurate classifiers for hyperspectral images?
➢SVMs perform a nonlinear pixel-wise classification based on the full spectral information which is robust to the spectral dimension of hyperspectral images.
The Hseg Segmentation The HSeg algorithm is a segmentation technique combining region growing, using the hierarchical stepwise optimization (HSWO) mIneittihaolidza, twiohni:chInpitrioadliuzceetshsepsaetgiamlleynctaotnionnecbteyd raesgsiigonnisn,gweitahchunpsiuxpeleravriesegdiocnlalassbiefli.caIftiaon, that gprroesuepgsmtoegnetathtieornsiismpilraorvsidpeadti,allalybedlisejaocihntpixel raecgcioorndsin. Tglhye. Oaltghoerriwthimse,claanbebleesaucmhmpiaxreizleads aas fsoepllaorwatse. region.
Abstract
Recent advances in spectral–spatial classification of hyperspectral images are presented in this paper. Several techniques are investigated for combining both spatial and spectral information.
But,it also brings some problem: the Curse of ➢DiTmheenfsiinoenaspliatytiaalnrdesthoeluntieoendofforthsepesecinfsicorsspeencatrballe–s sptahteiaal ncalalysssiifsieorfss. mall spatial structures in the image. ➢Many operational imaging systems are currently
➢Iterative statistical classifier based on Markov random field (MRF) modeling. Note that recently adaptive MRF have been introduced in remote sensing.
Highlight the importance of spectral–spatial strategies for the accurate classification of hyperspectral images and validate the proposed methods.
Introduction
PS: ➢Use advanced morphological filters as an 鲁棒alt性ern(Ratoivbeuswt)a:y即o系f p统er的for健mi壮ng性jo，in是t c在las异sif常ica和tio危n.
Introduction
2 PART TThWe HOseg Segmentation.
➢In high-dimensional spaces, normally distributed data have a tendency to concentrate in the tails, which seems to be contradictory with its bellshaped density function.
维数灾难(Curse of Dimensionality):通常是指在涉
Introduction
The Curse of Dimensionality of hyperspectral remote sensor technology :
Advantages of hyperspectral remote sensor technology:
➢The detailed spectral information increases the possibility of more accurately discriminating materials of interest.
➢the rate of convergence of the statistical estimation decreases when the dimension grows while conjointly the number of parameters to estimate increases, making the estimation of the model parameters very difficult.
Hierarchical SegmHeSnegtation
Reading Report
CONTENTS
PART ONE
PARTБайду номын сангаасTWO
PART THREE
About the Introduction. The Hseg Segmentation. Application in ENVI.
1 PART OAbNouEt the Introduction.
➢with a limited training set, beyond a certain limit,
Introduction
How to build accurate classifiers for hyperspectral images?
➢SVMs perform a nonlinear pixel-wise classification based on the full spectral information which is robust to the spectral dimension of hyperspectral images.
The Hseg Segmentation The HSeg algorithm is a segmentation technique combining region growing, using the hierarchical stepwise optimization (HSWO) mIneittihaolidza, twiohni:chInpitrioadliuzceetshsepsaetgiamlleynctaotnionnecbteyd raesgsiigonnisn,gweitahchunpsiuxpeleravriesegdiocnlalassbiefli.caIftiaon, that gprroesuepgsmtoegnetathtieornsiismpilraorvsidpeadti,allalybedlisejaocihntpixel raecgcioorndsin. Tglhye. Oaltghoerriwthimse,claanbebleesaucmhmpiaxreizleads aas fsoepllaorwatse. region.
Abstract
Recent advances in spectral–spatial classification of hyperspectral images are presented in this paper. Several techniques are investigated for combining both spatial and spectral information.
But,it also brings some problem: the Curse of ➢DiTmheenfsiinoenaspliatytiaalnrdesthoeluntieoendofforthsepesecinfsicorsspeencatrballe–s sptahteiaal ncalalysssiifsieorfss. mall spatial structures in the image. ➢Many operational imaging systems are currently
➢Iterative statistical classifier based on Markov random field (MRF) modeling. Note that recently adaptive MRF have been introduced in remote sensing.
Highlight the importance of spectral–spatial strategies for the accurate classification of hyperspectral images and validate the proposed methods.
Introduction
PS: ➢Use advanced morphological filters as an 鲁棒alt性ern(Ratoivbeuswt)a:y即o系f p统er的for健mi壮ng性jo，in是t c在las异sif常ica和tio危n.
Introduction
2 PART TThWe HOseg Segmentation.
➢In high-dimensional spaces, normally distributed data have a tendency to concentrate in the tails, which seems to be contradictory with its bellshaped density function.