Point cloud optimization method of low-altitude remote sensing image based on vertical patch-based least square matching

Recently, the demand for remote sensing image retrieval is growing and attracting the interest of many researchers because of the increasing number of remote sensing images. Hashing, as a method of retrieving images, has been widely applied to remote sensing image retrieval. In order to improve hashing performance, we develop a cohesion intensive deep hashing model for remote sensing image retrieval. The underlying architecture of our deep model is motivated by the state-of-the-art residual net. Residual nets aim at avoiding gradient vanishing and gradient explosion when the net reaches a certain depth. However, different from the residual net which outputs multiple class-labels, we present a residual hash net that is terminated by a Heaviside-like function for binarizing remote sensing images. In this scenario, the representational power of the residual net architecture is exploited to establish an end-to-end deep hashing model. The residual hash net is trained subject to a weighted loss strategy that intensifies the cohesiveness of image hash codes within one class. This effectively addresses the data imbalance problem normally arising in remote sensing image retrieval tasks. Furthermore, we adopted a gradualness optimization method for obtaining optimal model parameters in order to favor accurate binary codes with little quantization error. We conduct comparative experiments on large-scale remote sensing data sets such as UCMerced and AID. The experimental results validate the hypothesis that our method improves the performance of current remote sensing image retrieval.

Download Full-text

A Multi-View Dense Point Cloud Generation Algorithm Based on Low-Altitude Remote Sensing Images

Remote Sensing ◽

10.3390/rs8050381 ◽

2016 ◽

Vol 8 (5) ◽

pp. 381 ◽

Cited By ~ 12

Author(s):

Zhenfeng Shao ◽

Nan Yang ◽

Xiongwu Xiao ◽

Lei Zhang ◽

Zhe Peng

Keyword(s):

Remote Sensing ◽

Point Cloud ◽

Remote Sensing Images ◽

Generation Algorithm ◽

Dense Point ◽

Low Altitude

Download Full-text

Natural Resource Monitoring ofRheum tanguticumby Multilevel Remote Sensing

Evidence-based Complementary and Alternative Medicine ◽

10.1155/2014/618902 ◽

2014 ◽

Vol 2014 ◽

pp. 1-9 ◽

Cited By ~ 1

Author(s):

Caixiang Xie ◽

Jingyuan Song ◽

Fengmei Suo ◽

Xiwen Li ◽

Ying Li ◽

...

Keyword(s):

Remote Sensing ◽

Medicinal Plants ◽

Remote Sensing Image ◽

Vegetation Coverage ◽

Resource Monitoring ◽

Satellite Imaging ◽

Ground Investigation ◽

Natural Resource Monitoring ◽

Low Altitude ◽

Rheum Tanguticum

Remote sensing has been extensively applied in agriculture for its objectiveness and promptness. However, few applications are available for monitoring natural medicinal plants. In the paper, a multilevel monitoring system, which includes satellite and aerial remote sensing, as well as ground investigation, was initially proposed to monitor naturalRheum tanguticumresource in Baihe Pasture, Zoige County, Sichuan Province. The amount ofR. tanguticumfrom images isM=S*ρandSis vegetation coverage obtained by satellite imaging, whereasρisR. tanguticumdensity obtained by low-altitude imaging. Only theR. tanguticumwhich coverages exceeded 1 m2could be recognized from the remote sensing image because of the 0.1 m resolution of the remote sensing image (called effective resource at that moment), and the results of ground investigation represented the amounts ofR. tanguticumresource in all sizes (called the future resource). The data in paper showed that the present available amount ofR. tanguticumaccounted for 4% to 5% of the total quantity. The quantity information and the population structure ofR. tanguticumin the Baihe Pasture were initially confirmed by this system. It is feasible to monitor the quantitative distribution for natural medicinal plants with scattered distribution.

Download Full-text