{"title":"Surface Depth Estimation From Multiview Stereo Satellite Images With Distribution Contrast Network","authors":"Ziyang Chen;Wenting Li;Zhongwei Cui;Yongjun Zhang","doi":"10.1109/JSTARS.2024.3457616","DOIUrl":null,"url":null,"abstract":"The calculation of surface depth based on multiview \n<bold>s</b>\ntereo (MVS) satellite imagery is of significant importance in fields such as military and surveying. The challenge in extracting depth information from satellite imagery lies in the fact that these images often exhibit similar colors, necessitating the development of algorithms that can integrate shape and texture information. Moreover, the application of classical convolutional neural network (CNN) MVS is limited by its inability to capture long-range terrain relationships, which presents a bottleneck in existing surface depth estimation algorithms. To address the above problems, we propose the Distribution Contrast Network for Surface Depth Estimation from Satellite Multi\n<bold>V</b>\niew \n<bold>S</b>\ntereo Images (DC-SatMVS), a novel satellite MVS network. In order to learn short-range and long-range features, we designed separate CNN and ViT branches. To emphasize the importance of shape and texture, we propose the Distribution Contrast Loss mechanism. This mechanism supervises the model training based on the similarity between the predicted depth and the ground truth depth distribution. Experimental results demonstrate that our method achieves state-of-the-art (SOTA) performance. We produce a remarkable 18.14% reduction in root mean square error compared to the Sat-MVSF on the WHU-TLC dataset. To validate the generalization performance of our framework, we trained and tested it on the DTU dataset, a common MVS dataset, and achieve SOTA results in this dataset as well.","PeriodicalId":13116,"journal":{"name":"IEEE Journal of Selected Topics in Applied Earth Observations and Remote Sensing","volume":null,"pages":null},"PeriodicalIF":4.7000,"publicationDate":"2024-09-24","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"https://ieeexplore.ieee.org/stamp/stamp.jsp?tp=&arnumber=10689488","citationCount":"0","resultStr":null,"platform":"Semanticscholar","paperid":null,"PeriodicalName":"IEEE Journal of Selected Topics in Applied Earth Observations and Remote Sensing","FirstCategoryId":"5","ListUrlMain":"https://ieeexplore.ieee.org/document/10689488/","RegionNum":2,"RegionCategory":"地球科学","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":null,"EPubDate":"","PubModel":"","JCR":"Q1","JCRName":"ENGINEERING, ELECTRICAL & ELECTRONIC","Score":null,"Total":0}
引用次数: 0
Abstract
The calculation of surface depth based on multiview
s
tereo (MVS) satellite imagery is of significant importance in fields such as military and surveying. The challenge in extracting depth information from satellite imagery lies in the fact that these images often exhibit similar colors, necessitating the development of algorithms that can integrate shape and texture information. Moreover, the application of classical convolutional neural network (CNN) MVS is limited by its inability to capture long-range terrain relationships, which presents a bottleneck in existing surface depth estimation algorithms. To address the above problems, we propose the Distribution Contrast Network for Surface Depth Estimation from Satellite Multi
V
iew
S
tereo Images (DC-SatMVS), a novel satellite MVS network. In order to learn short-range and long-range features, we designed separate CNN and ViT branches. To emphasize the importance of shape and texture, we propose the Distribution Contrast Loss mechanism. This mechanism supervises the model training based on the similarity between the predicted depth and the ground truth depth distribution. Experimental results demonstrate that our method achieves state-of-the-art (SOTA) performance. We produce a remarkable 18.14% reduction in root mean square error compared to the Sat-MVSF on the WHU-TLC dataset. To validate the generalization performance of our framework, we trained and tested it on the DTU dataset, a common MVS dataset, and achieve SOTA results in this dataset as well.
期刊介绍:
The IEEE Journal of Selected Topics in Applied Earth Observations and Remote Sensing addresses the growing field of applications in Earth observations and remote sensing, and also provides a venue for the rapidly expanding special issues that are being sponsored by the IEEE Geosciences and Remote Sensing Society. The journal draws upon the experience of the highly successful “IEEE Transactions on Geoscience and Remote Sensing” and provide a complementary medium for the wide range of topics in applied earth observations. The ‘Applications’ areas encompasses the societal benefit areas of the Global Earth Observations Systems of Systems (GEOSS) program. Through deliberations over two years, ministers from 50 countries agreed to identify nine areas where Earth observation could positively impact the quality of life and health of their respective countries. Some of these are areas not traditionally addressed in the IEEE context. These include biodiversity, health and climate. Yet it is the skill sets of IEEE members, in areas such as observations, communications, computers, signal processing, standards and ocean engineering, that form the technical underpinnings of GEOSS. Thus, the Journal attracts a broad range of interests that serves both present members in new ways and expands the IEEE visibility into new areas.