Results 11 to 20 of about 11,124 (155)
Integrative Conservation, Volume 2, Issue 2, Page 80-94, June 2023., 2023 The objective of this study is to characterise ecological connectivity for mammals and identify important patches and linkages for connecting urban green spaces for Greater Kuala Lumpur, Malaysia. Our results demonstrate the potential for Southeast Asian megacities to support biodiversity in the urban context, and the potential for a different kind of ...
Jennifer Danneck +6 more
wiley +1 more source
STDC‐MA network for semantic segmentation
IET Image Processing, Volume 16, Issue 14, Page 3758-3767, 11 December 2022., 2022 Abstract Semantic segmentation is applied extensively in autonomous driving and intelligent transportation with methods that highly demand spatial and semantic information. Here, an STDC‐MA network is proposed to meet these demands. First, the STDC‐Seg structure is employed in STDC‐MA to ensure a lightweight and efficient structure.
Xiaochun Lei +6 more
wiley +1 more source
Dark feet and dark wings: penetrating the depths of the Earth
Journal of Analytical Psychology, Volume 67, Issue 5, Page 1341-1362, November 2022., 2022 Abstract The author, although an analyst, is an initiate into the topic of environmental destruction. Following Wendell Berry, she enters the dark and begins a journey of dream‐like reflection, weaving images from her own dream and drawing on the work of Vaughan, Bernstein, Soloveitchik and Sacks.
Robin B. Zeiger
wiley +1 more source
Cascading and Residual Connected Network for Single Image Superresolution
Wireless Communications and Mobile Computing, Volume 2021, Issue 1, 2021., 2021 Convolution neural networks facilitate the significant process of single image super‐resolution (SISR). However, most of the existing CNN‐based models suffer from numerous parameters and excessively deeper structures. Moreover, these models relying on in‐depth features commonly ignore the hints of low‐level features, resulting in poor performance. This
Kai Huang +6 more
wiley +1 more source
Discriminative Codebook Hashing for Supervised Video Retrieval
Computational Intelligence and Neuroscience, Volume 2021, Issue 1, 2021., 2021 In recent years, hashing learning has received increasing attention in supervised video retrieval. However, most existing supervised video hashing approaches design hash functions based on pairwise similarity or triple relationships and focus on local information, which results in low retrieval accuracy.
Xiaoman Bian +7 more
wiley +1 more source
Image enhancement method based on exposure fusion for UAV aerial photography
Xibei Gongye Daxue Xuebao, 2022 针对无人机航拍图像光照不均匀及自然雾导致影像质量退化问题, 提出了一种无人机航拍图像增强算法。利用改进的低照度图像增强算法均衡亮度对比度; 为了解决均衡后图像过增强问题, 提出了联合去雾及曝光融合的色彩矫正增强方法; 为了保留增强图像的边缘纹理信息, 设计了一种效果更佳的细节增强算法, 处理后统计直方图更为平滑, 可在一定程度上抑制部分噪声, 细节纹理信息更强。实验结果表明, 所提的航拍图像增强算法, 能够有效解决因光照不均或自然雾引起的影像退化现象, 提高了无人机航拍图像的质量 ...
LI Liangliang +6 more
doaj +1 more source
Journal of Mathematics, Volume 2021, Issue 1, 2021., 2021 Fractional order research has interdisciplinary characteristics and has been widely used in the field of natural sciences. Therefore, fractional order research has become an important area of concern for scholars. This paper used 2854 literatures collected from China National Knowledge Infrastructure (CNKI) database from 2001 to 2020 as the data source
Yunfei Yang +4 more
wiley +1 more source
Zhejiang Daxue xuebao. Lixue ban, 2005 介绍了一个采用自顶向下、逐步细分的结构化软件设计方法，以Visual C++6.0为软件开发平台设计并实现的自主版权的大数据量遥感图像数字处理系统的算法模块的设计、结构，以及系统实现中采用的关键技术.提出了美国ERDAS图像处理系统在滤波算法的图像边界处理及图像处理后象元值超限处理两方面的不足，同时提出并实现了新的优化方法.实践证明在这两方面的优化更有利于图像分析和信息提取.
JINGChang-feng(靖常峰) +2 more
doaj +1 more source
Control Performance Improvement of Cutting and Indentation Equipment in Paper Packing with Machine Vision Technology(机器视觉技术对纸质包装切割压痕设备控制性能的改进设计) [PDF]
Shipin kexue jishu xuebao, 2014 Due to problems of NRG-P type cutting and indentation equipment, a new test method, the machine vision technology, was presented to improve its control performance.
HUO Liang-sheng(霍亮生) +2 more
doaj +1 more source
全球能源互联网, 2019 电力巡检无人机提供大量的巡检图像，但由于电力设备故障稀少，其中只有少量的设备缺陷故障图像可供使用。为了提高设备缺陷故障识别精度并减少训练过拟合问题，首先介绍了基于迁移学习的小样本电力巡检图像处理方法，通过图像裁剪、翻转、旋转等数据增强技术对小样本图像进行扩充，同时采用生成对抗网络（GAN）来扩充基础样本；并使用迁移学习技术，将基于大规模图像数据的预训练深度卷积模型进行定制，调整该神经网络模型的输入层和最后两层参数，并对超参进行调优。实验结果表明，巡检设备故障（如导线断股和绝缘子串脱落等）准确匹配度近95%
陆继翔 +4 more
doaj +1 more source