Object detection based on deep learning has achieved promising results on traditional datasets, but detecting objects in remotesensingimagery with diverse occlusions remains challenging. This is due to the fact that...
详细信息
ISBN:
(纸本)9783031451690;9783031451706
Object detection based on deep learning has achieved promising results on traditional datasets, but detecting objects in remotesensingimagery with diverse occlusions remains challenging. This is due to the fact that natural occlusions are common in real-world images, and there is a lack of adequate datasets and neglect of latent information that can be useful for identification. Our research endeavors to tackle this challenge by presenting an innovative single-stage image-adaptive YOLO-transformer framework that leverages attention-based mechanisms to enable the model to concentrate on important regions and extract more distinctive features. To optimize the model's accuracy while maintaining its lightweight and suitability for real-time applications, we employed a depthwise convolution and SiLU activation function in lieu of the standard convolution. These modifications allow our framework to attain high levels of precision. Our approach is evaluated on synthetically generated occluded datasets from publicly available NWPU-VHR10 and demonstrated adaptive processing of images in both normal and three types of environmental occlusions (foggy, rainy, and cloudy). The experimental results are highly promising, as our approach achieved an inference speed of 6 ms with 11.9 GFLOPs on the NVIDIA Tesla T4 while maintaining effectiveness in all three occlusion types.
David Landgrebe, Professor of Electrical Engineering and Director of the Purdue University Laboratory for Applications of remotesensing (LARS), was a primary innovator in the field of digital image analysis and remot...
详细信息
David Landgrebe, Professor of Electrical Engineering and Director of the Purdue University Laboratory for Applications of remotesensing (LARS), was a primary innovator in the field of digital image analysis and remotesensing of the environment. He and his LARS colleagues, along with a selected few other researchers at institutions including the University of Michigan, University of California Berkley, NASA Goddard Space Flight Center, and NASA Johnson Space Center, defined and developed remotesensing technology to monitor the Earth's terrestrial environment. This research led to the Landsat program, which has continued to monitor the Earth's land areas for a half century. These technologies have defined new fields of scientific query in digital image analysis, biophysical remotesensing, as well as remotesensing science and applications. Dr. Landgrebe's contributions to these research areas were substantial and profound. Understanding the early evolution of work is critical to understanding how this technology is still advancing today. The authors hope that current and future students of these fields will benefit from understanding how this all began.
remotesensingimages have a wide range of applications in geological exploration, disaster warning, military reconnaissance and other fields, and the detection of specific targets in remotesensingimages can improve...
详细信息
The advancement of Earth observation technologies has enabled comprehensive Earth monitoring at unprecedented levels. Single remotesensingimages often fall short in practical applications due to various limitations....
详细信息
The image classification as the key technology in the remotesensing system is mainly based on the image with the characteristics of the electromagnetic wave radiation information of the ground object to distinguish a...
详细信息
To a certain extent, foreign objects on road surfaces can cause damage to vehicles and increase the risk of accidents, which is even more immeasurably harmful in such special scenarios as airports. Effective detection...
详细信息
In practical applications, remotesensing (RS) scene classification faces data shift problems, including novel class and data discrepancy problems. Due to these problems, it is difficult to obtain representative and d...
详细信息
Nowadays, aircraft positioning is mainly based on GNSS, but GNSS signal receiving equipment is difficult to ensure the stability and reliability of the positioning signal in the presence of obstruction or strong elect...
详细信息
This paper focuses on super-resolution of natural scene text images, serving as a crucial preprocessing step for imagerecognition, with the aim of extracting text information from low-resolution images. In this prepr...
详细信息
Since the temporal variation for growing period of paddy rice can be shown clearly in optical and radar images, a Long-Short Term Memory (LSTM) model is introduced to construct paddy rice recognition systems using Sen...
详细信息
暂无评论