Deep capsule network combined with spatial-spectral information for hyperspectral image classification

نویسندگان

چکیده

å‘é‡åŒ–çš„èƒ¶å›Šç¥žç»å ƒå’ŒåŠ¨æ€è·¯ç”±å¼çš„ä¿¡æ¯ä¼ é€’æœºåˆ¶èµ‹äºˆäº†èƒ¶å›Šç½‘ç»œæ›´å¼ºçš„ç‰¹å¾è¡¨ç¤ºèƒ½åŠ›ã€‚åœ¨é¥æ„Ÿé¢†åŸŸï¼ŒåŸºäºŽèƒ¶å›Šç½‘ç»œçš„é«˜å ‰è°±å½±åƒåˆ†ç±»æ–¹æ³•å·²ç»èŽ·å¾—äº†è¾ƒä¼ ç»Ÿæ·±åº¦å­¦ä¹ æ¨¡åž‹æ›´ä¸ºä¼˜å¼‚çš„åˆ†ç±»ç»“æžœã€‚é’ˆå¯¹çŽ°æœ‰èƒ¶å›Šåˆ†ç±»æ¨¡åž‹ä¸­å­˜åœ¨çš„ç½‘ç»œæµ å±‚ã€ç©ºè°±è”åˆä¿¡æ¯åˆ©ç”¨ä¸è¶³ç­‰é—®é¢˜ï¼Œæœ¬æ–‡åˆ©ç”¨å·ç§¯èƒ¶å›Šå±‚ã€æ®‹å·®è¿žæŽ¥ã€ä¸‰ç»´å·ç§¯èƒ¶å›Šå±‚æž„å»ºäº†ä¸€ç§ç”¨äºŽé«˜å ‰è°±å½±åƒåˆ†ç±»çš„æ–°åž‹æ·±åº¦èƒ¶å›Šç½‘ç»œã€‚å ·ä½“åœ°ï¼Œæœ¬æ–‡æ–¹æ³•ç›´æŽ¥ä»¥é«˜ç»´æ•°æ®ç«‹æ–¹ä½“ä½œä¸ºç½‘ç»œè¾“å ¥ï¼Œå¹¶åˆ©ç”¨èƒ¶å›Šæ®‹å·®å—é€å±‚æå–æ•°æ®ä¸­çš„æ·±å±‚æŠ½è±¡ç‰¹å¾ã€‚ä¸ºäº†æ›´åŠ å åˆ†åœ°åˆ©ç”¨å½±åƒä¸­çš„ç©ºè°±è”åˆç‰¹å¾ï¼Œåœ¨æ·±å±‚æ¬¡çš„èƒ¶å›Šæ®‹å·®å—ä¸­å¼•å ¥ä¸‰ç»´å·ç§¯èƒ¶å›Šå±‚ï¼Œä»¥è¿›ä¸€æ­¥æé«˜åˆ†ç±»ç²¾åº¦ã€‚ä¸ºäº†éªŒè¯æœ¬æ–‡æ–¹æ³•çš„æœ‰æ•ˆæ€§ï¼Œé€‰æ‹©University of Pavia、Indian Pines和Salinas等3ä¸ªå¸¸ç”¨é«˜å ‰è°±æ•°æ®é›†å’Œä¸€ä¸ªå¤§è§„æ¨¡æœºè½½é«˜å ‰è°±æ•°æ®é›†Chikuseiè¿›è¡Œå®žéªŒã€‚ç»“æžœè¡¨æ˜Žï¼Œä¸ŽçŽ°æœ‰æ·±åº¦å­¦ä¹ æ¨¡åž‹ç›¸æ¯”ï¼Œæœ¬æ–‡æ–¹æ³•èƒ½å¤ŸèŽ·å¾—æ›´ä¸ºä¼˜å¼‚çš„åˆ†ç±»æ•ˆæžœï¼Œåœ¨4个数据集上分别获得了99.43%、98.85%、97.14%和97.43%的总体分类精度。

برای دانلود باید عضویت طلایی داشته باشید

برای دانلود متن کامل این مقاله و بیش از 32 میلیون مقاله دیگر ابتدا ثبت نام کنید

اگر عضو سایت هستید لطفا وارد حساب کاربری خود شوید

منابع مشابه

Spectral-Spatial Response for Hyperspectral Image Classification

This paper presents a hierarchical deep framework called Spectral-Spatial Response (SSR) to jointly learn spectral and spatial features of Hyperspectral Images (HSIs) by iteratively abstracting neighboring regions. SSR forms a deep architecture and is able to learn discriminative spectral-spatial features of the input HSI at different scales. It includes several existing spectral-spatial-based ...

متن کامل

Hyperspectral Image Classification Based on Nonlinear Spectral-Spatial Network

Recently, for the task of hyperspectral images classification, deep learning-based methods have revealed promising performance. However, the complex network structure and time-consuming training process have restricted their applications. In this letter, we construct a much simpler network, nonlinear spectral-spatial network (NSSNet), for hyperspectral images classification. NSSNet is developed...

متن کامل

Spectral-Spatial Hyperspectral Image Classification With Edge-Preserving Filtering

The integration of spatial context in the classification of hyperspectral images is known to be an effective way in improving classification accuracy. In this paper, a novel spectralspatial classification framework based on edge-preserving filtering is proposed. The proposed framework consists of the following three steps. First, the hyper-spectral image is classified using a pixel-wise classif...

متن کامل

A Diversified Deep Belief Network for Hyperspectral Image Classification

In recent years, researches in remote sensing demonstrated that deep architectures with multiple layers can potentially extract abstract and invariant features for better hyperspectral image classification. Since the usual real-world hyperspectral image classification task cannot provide enough training samples for a supervised deep model, such as convolutional neural networks (CNNs), this work...

متن کامل

Spectral/Spatial Hyperspectral Image Compression

^Remote Sensing Signal and Image Processing Laboratory Department of Computer Science and Electrical Engineering University of Maryland, Baltimore County, Baltimore, MD 21250 ^Computer Science Department, University of Extremadura Avda. de la Universidad s/n,10.071 Caceres, SPAIN ^Center for Space and Remote Sensing Research Graduate Institute of Space Science Department of Computer Science and...

متن کامل

ذخیره در منابع من


  با ذخیره ی این منبع در منابع من، دسترسی به آن را برای استفاده های بعدی آسان تر کنید

ژورنال

عنوان ژورنال: Journal of remote sensing

سال: 2021

ISSN: ['1007-4619', '2095-9494']

DOI: https://doi.org/10.11834/jrs.20210309