Classification of high-resolution remote sensing images based on enhanced DeepLab algorithm and adaptive loss function

نویسندگان

چکیده

é«˜åˆ†è¾¨çŽ‡é¥æ„Ÿå½±åƒåœ°ç‰©å¤æ‚ï¼Œåˆ†ç±»éš¾åº¦å¤§ï¼Œè€Œæ·±åº¦å­¦ä¹ æ–¹æ³•å¯ä»¥æå–åœ°ç‰©æ›´å¤šæ›´æ·±å±‚æ¬¡çš„ç‰¹å¾ä¿¡æ¯ï¼Œé€‚ç”¨äºŽé«˜åˆ†è¾¨çŽ‡é¥æ„Ÿå½±åƒçš„åœ°ç‰©åˆ†ç±»ã€‚æœ¬æ–‡ç ”ç©¶å¯¹é«˜åˆ†è¾¨çŽ‡å½±åƒä¸­ä¸é€æ°´åœ°é¢ã€å»ºç­‘ã€ä½ŽçŸ®æ¤è¢«ã€æ ‘ã€è½¦è¾†ç­‰åœ°ç‰©çš„é«˜ç²¾åº¦åˆ†ç±»ã€‚ç»“åˆé¥æ„Ÿå¤šåœ°ç‰©åˆ†ç±»çš„ç‰¹ç‚¹ï¼Œä»¥DeepLab v3+网络模型为基础,提出E-DeepLab网络模型。主要改进为:(1ï¼‰æ”¹è¿›ç¼–ç å™¨å’Œè§£ç å™¨çš„ç»“åˆæ–¹å¼ï¼Œä½¿ç”¨ç®€æ´æœ‰æ•ˆçš„åŠ æˆè¿žæŽ¥æ–¹å¼ã€‚ï¼ˆ2ï¼‰ç¼©å°å•æ¬¡ä¸Šé‡‡æ ·å€æ•°ï¼Œå¢žåŠ ä¸Šé‡‡æ ·å±‚ï¼Œæé«˜ç¼–ç å™¨ä¸Žè§£ç å™¨è¿žæŽ¥çš„ç´§å¯†æ€§ã€‚ï¼ˆ3ï¼‰ä½¿ç”¨æ”¹è¿›çš„è‡ªé€‚åº”æƒé‡æŸå¤±å‡½æ•°ï¼Œè‡ªåŠ¨è°ƒèŠ‚åœ°ç‰©æŸå¤±æƒé‡ã€‚åŒæ—¶æ ¹æ®æ•°æ®ç‰¹ç‚¹ï¼Œæå‡ºç»“åˆDSM、NDVI数据等多通道训练方式。使用两个地区数据进行实验,结果表明,两地区精度均明显优于原始DeepLab v3+æ¨¡åž‹å’Œå ¶ä»–ç›¸å ³æ¨¡åž‹ï¼ŒPotsdam地区总体提取精度达到93.2%,建筑物提取精度达到97.8%,Vaihingen地区总体提取精度达到90.7%,建筑物提取精度达到96.3%ã€‚ç›®è§†å¯¹æ¯”åˆ†ç±»å›¾å’Œæ ‡å‡†æ ‡è®°å›¾ï¼Œä¸¤è€ å ·æœ‰é«˜åº¦çš„ä¸€è‡´æ€§ã€‚æœ¬æ–‡æ‰€æå‡ºçš„E-DeepLab网络在高分辨率遥感影像地物高精度提取和分类中有较好的应用价值。

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

Segmentation Improvement of High Resolution Remote Sensing Images based on superpixels using Edge-based SLIC algorithm (E-SLIC)

The segmentation of high resolution remote sensing images is one of the most important analyses that play a significant role in the maximal and exact extraction of information.  There are different types of segmentation methods among which using  superpixels is one of the most important ones. Several methods have been proposed for extracting superpixels. Among the most successful ones, we can r...

متن کامل

High Efficient Classification on Remote Sensing Images Based on Svm

Automatic image categorization using low-level features is a challenging research topic in remote sensing application. In this paper, we formulate the image categorization problem as an image texture learning problem by viewing an image as a collection of regions, each obtained from image segmentation. Our approach performs an effective feature mapping through a chosen metric distance function....

متن کامل

Farmland Parcels Extraction Based on High Resolution Remote Sensing Images

Extracting farmland parcels from high resolution remote sensing images is an important issue for land-use dynamic monitoring, precision agriculture and other fields. However, the traditional method, using GIS software and manual digital, has wasted a lot of human and material resources. In addition, the results are impacted by the human factors obviously. Therefore, an automatically extraction ...

متن کامل

Edge Fining Algorithm of Texture Classification for High Resolution Remote Sensing Images

In order to improve the classification accuracy of the high resolution remote sensing images, the algorithm based on texture features is proposed in this paper. The image is classified initially by texture features which are derived from gray level co-occurrence matrices. And the suitable weights of these features are given to form the feature vector. Secondly, the edge fining method, which can...

متن کامل

Graph-based semi-supervised classification on very high resolution remote sensing images

Classification of very high resolution (VHR) remote sensing imagery is a rapidly emerging discipline but faces several challenges owing to the huge scale of the pixel data involved, indiscernibility in the traditionally used features to represent various regions, and the lack of available ground truth data. This paper provides a framework which elegantly overcomes these hurdles by providing a n...

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: 2022

ISSN: ['1007-4619', '2095-9494']

DOI: https://doi.org/10.11834/jrs.20209200