International Research Journal of Engineering and Technology (IRJET)
e-ISSN: 2395-0056
Volume: 11 Issue: 08 | Aug 2024
p-ISSN: 2395-0072
www.irjet.net
Saliency segmentation and Structure LBP feature model for ship detection from satellite Images Aziha M 1, Dr. Jasmine J.C Sheeja, M.E., Ph.D., 2 1PG Student, Dept. of Electronics And Communication, Rohini College of Engineering and Technology, Kanniyakumari
Tamil Nadu, India 2Professor, Dept. of Electronics And Communication, Rohini College of Engineering and Technology,
Kanniyakumari, Tamil Nadu, India. -----------------------------------------------------------------------------***----------------------------------------------------------------------------However, high-level speckles are common in SAR images, Abstract — The acquisition of satellite images with the they are insensitive to wood, and they are difficult for humans to interpret. Optical satellite images have a higher resolution and contain more detailed information than SAR and other types of remote sensing images; Consequently, they are better suited for target recognition or detection.
highest resolutions is not possible because of the current optical satellites' technical limitations as well as their limited budgets. In existing framework, produce items with high ghastly (HS) and fleeting goals, presented a twostream otherworldly worldly combination method in view of consideration component called STA-Net. Because of the intricate attention mechanisms that are used both spatially and temporally, STA-Nets frequently necessitate a significant amount of computational power. This project proposes a detection algorithm based on ship structure and the local binary pattern (LBP) descriptor to overcome the shortcomings. Due to cluttered scenes and varying ship sizes, ship detection from optical satellite imagery is difficult. Present a novel saliency segmentation framework that allows for the flexible integration of multiple visual cues to extract candidate regions from various sea surfaces. This is important for a variety of applications, including illegal smuggling, traffic surveillance, fishery management, and so on. Then, straightforward shape analysis is used to get rid of clearly false targets. Finally, true ship targets are distinguished using a structure-LBP feature that identifies ships' inherent topology structure. Multiple panchromatic satellite image results confirm that the proposed method outperforms other current methods in terms of detection time and accuracy.
Yuan Yao An et al. [1] proposed In the field of remote sensing, ship target detection in optical images has received increasing attention. Optical remote sensing images' ship target detection technology is susceptible to numerous factors, whereas real data are difficult to contain. To get the different circumstances in the huge ocean scenes, we foster a reproduction framework for high-goal optical remote detecting picture of boat targets. Xiaoyang Xie et al. [2] proposed Ship distribution probability analysis enables rapid ship detection from optical satellite images. It is still difficult to detect ships automatically using optical satellite images. By analyzing the ship distribution probability, this paper proposes a novel method for ship detection from optical satellites. The sea cluster histogram model is used to first construct an anomaly detection model; The ship candidates are then identified by examining the ship distribution in light of the ship safety navigational criterion, and the area properties of the ship candidates remove obvious non-ship objects; Lastly, a ship candidate's structural continuity descriptor is intended to eliminate false alarms.
Key Words: Hyperspectral image, Local Binary Pattern (LBP), Saliency Segmentation, Shape Analysis.
Shuchen Wang et al. [3] proposed The Saliency Adjusted Deep Network for Optical Satellite Image Ship Detection. a Saliency Adjusted YOLO (SA-YOLO) for optical satellite image ship detection is developed. First, due to the fact that the ship in low resolution imagery can be regarded as a salient object, they designed a saliency guided dense sampling layer (SDSL) to improve the spatial sampling of small ship targets. Secondly, the saliency region-aware convolution (SAConv) strategy is designed to improve the representation capability of salient regions and increase the attention of network to these regions. Sergey Voinov et al.
1. INTRODUCTION For a wide range of applications, such as traffic surveillance, fishery management, and illegal smuggling, remote sensing imagery's ability to identify ships is crucial. Because they are little affected by weather and time, synthetic aperture radar (SAR) images play an important role in detecting and tracing targets in previous research.
© 2024, IRJET
|
Impact Factor value: 8.226
|
ISO 9001:2008 Certified Journal
|
Page 54