Optical remote sensing image object detection based on multi-resolution feature fusion

نویسندگان

چکیده

é«˜åˆ†è¾¨çŽ‡é¥æ„Ÿå›¾åƒç›®æ ‡æ£€æµ‹æ˜¯è®¡ç®—æœºè§†è§‰çš„ä¸€ä¸ªé‡è¦ç ”ç©¶é¢†åŸŸï¼Œåœ¨æ°‘ç”¨ä¸Žå†›äº‹é¢†åŸŸå ·æœ‰é‡è¦çš„åº”ç”¨ä»·å€¼ã€‚ç›®å‰ï¼ŒåŸºäºŽæ·±åº¦å­¦ä¹ çš„è‡ªç„¶å›¾åƒç›®æ ‡æ£€æµ‹æœ‰äº†çªç ´æ€§è¿›å±•ã€‚ä½†æ˜¯ï¼Œç”±äºŽé¥æ„Ÿå›¾åƒå ·æœ‰ç›®æ ‡å°ºåº¦å·®å¼‚å¤§ä¸”ç±»é—´ç›¸ä¼¼åº¦é«˜çš„ç‰¹ç‚¹ï¼Œä½¿å¾—å¤„ç†è‡ªç„¶å›¾åƒçš„ç›®æ ‡æ£€æµ‹ç®—æ³•ç›´æŽ¥åº”ç”¨äºŽé¥æ„Ÿå›¾åƒæ—¶ä»é¢ä¸´ç€ä¸€äº›æŒ‘æˆ˜ã€‚é’ˆå¯¹ä¸Šè¿°æŒ‘æˆ˜ï¼Œæœ¬æ–‡æå‡ºä¸€ç§å¤šåˆ†è¾¨çŽ‡ç‰¹å¾èžåˆçš„é¥æ„Ÿå›¾åƒç›®æ ‡æ£€æµ‹æ–¹æ³•ã€‚é¦–å ˆï¼Œé€šè¿‡ç‰¹å¾é‡‘å­—å¡”æå–å¤šå°ºåº¦ç‰¹å¾å›¾å¹¶åœ¨å ¶åŽåµŒå ¥å¤šåˆ†è¾¨çŽ‡ç‰¹å¾æå–ç½‘ç»œï¼Œä¿ƒä½¿ç½‘ç»œå­¦ä¹ ç›®æ ‡åœ¨ä¸åŒåˆ†è¾¨çŽ‡ä¸‹çš„ç‰¹å¾ï¼Œç¼©å°ä¸åŒç‰¹å¾å±‚ä¹‹é—´çš„è¯­ä¹‰å·®è·ã€‚å ¶æ¬¡ï¼Œä¸ºå®žçŽ°å¤šåˆ†è¾¨ç‰¹å¾çš„æœ‰æ•ˆèžåˆï¼Œæœ¬æ–‡é‡‡ç”¨è‡ªé€‚åº”ç‰¹å¾èžåˆæ¨¡å—æŒ–æŽ˜æ›´å ·åˆ¤åˆ«æ€§çš„å¤šåˆ†è¾¨ç‰¹å¾è¡¨è¾¾ã€‚æœ€åŽï¼Œå°†è‡ªé€‚åº”ç‰¹å¾èžåˆæ¨¡å—çš„è¾“å‡ºç‰¹å¾çš„ç›¸é‚»å±‚è¿›è¡Œæ·±åº¦èžåˆã€‚åœ¨å ¬å¼€çš„é¥æ„Ÿå›¾åƒç›®æ ‡æ£€æµ‹æ•°æ®é›†DIOR和DOTA上评估了本文方法的有效性,相比采用特征金字塔结构的Faster R-CNN,本文方法的准确率(mAP)分别提高2.5%和2.2%。

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

A detection method of artificial area from high resolution remote sensing images based on multi scale and multi feature fusion

In order to solve the problem of automatic detection of artificial objects in high resolution remote sensing images, a method for detection of artificial areas in high resolution remote sensing images based on multi-scale and multi feature fusion is proposed. Firstly, the geometric features such as corner, straight line and right angle are extracted from the original resolution, and the pseudo ...

متن کامل

Object-oriented change detection approach for high-resolution remote sensing images based on multiscale fusion

Aiming at the difficulties in change detection caused by the complexity of highresolution remote sensing images that exist in varied ecological environments and artificial objects, in order to overcome the limitations in traditional pixel-oriented change detection methods and improve the detection precision, an innovative object-oriented change detection approach based on multiscale fusion is p...

متن کامل

High Resolution Remote Sensing Image Segmentation Based on Multi-features Fusion

Article history: Received: 7.1.2016. Received in revised form: 6.5.2016. Accepted: 7.5.2016. High resolution remote sensing images contain richer information of spatial relation in ground objects than low resolution ones, which can help describe the geometric information and extract the essential features more efficiently. However, the handling difficulties due to the relative poorer spectral i...

متن کامل

A Study of High-resolution Remote Sensing Image Data Fusion Based on Multi-level Techniques

Multi-source remote sensing image fusion was largely focused on in the recent years, especially in which high-resolution remote sensing image data fusion (HRRSIDF) was one of its important centers and is going into the main flow of the future remote sensing image data fusion. In the paper, the authors principally illustrated a concept model, constructed the RSIDF’s (Remote Sensing Image Data Fu...

متن کامل

Object-oriented Change Detection for Remote Sensing Images Based on Multi-scale Fusion

In the process of object-oriented change detection, the determination of the optimal segmentation scale is directly related to the subsequent change information extraction and analysis. Aiming at this problem, this paper presents a novel object-level change detection method based on multi-scale segmentation and fusion. First of all, the fine to coarse segmentation is used to obtain initial obje...

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.20210505