Dataset generation and validation for spacecraft pose estimation via monocular images processing

Bechini, Michele; Lavagna, Michèle; Lunghi, Paolo

doi:10.1016/j.actaastro.2023.01.012

Cited by 13 publications

(11 citation statements)

References 17 publications

Supporting

Mentioning

Contrasting

Order By: Relevance

“…Real satellite component data sets are costly, but the good news is that some researchers have broken the limitation of data set scarcity to some extent by synthesizing realistic data sets. These data sets can be used for pose estimation (Kisantal et al , 2020; Bechini et al , 2022) and even instance segmentation (Faraco et al , 2022; Proença and Gao, 2020). Considering the availability of depth information, a public data set AFDL-SCD [1] is selected as the main testing data.…”

Section: Experiments and Analysismentioning

confidence: 99%

A satellite component contour extraction method for lightweight space mobile platforms

Zhu

Liang

2023

AEAT

View full text Add to dashboard Cite

Purpose Owing to the complex space environment and limited computing resources, traditional and deep learning-based methods cannot complete the task of satellite component contour extraction effectively. To this end, this paper aims to propose a high-quality real-time contour extraction method based on lightweight space mobile platforms. Design/methodology/approach A contour extraction method that combines two edge clues is proposed. First, Canny algorithm is improved to extract preliminary contours without inner edges from the depth images. Subsequently, a new type of edge pixel feature is designed based on surface normal. Finally, surface normal edges are extracted to supplement the integrity of the preliminary contours for contour extraction. Findings Extensive experiments show that this method can achieve a performance comparable to that of deep learning-based methods and can achieve a 36.5 FPS running rate on mobile processors. In addition, it exhibits better robustness under complex scenes. Practical implications The proposed method is expected to promote the deployment process of satellite component contour extraction tasks on lightweight space mobile platforms. Originality/value A pixel feature for edge detection is designed and combined with the improved Canny algorithm to achieve satellite component contour extraction. This study provides a new research idea for contour extraction and instance segmentation research.

show abstract

Section: Experiments and Analysismentioning

confidence: 99%

A satellite component contour extraction method for lightweight space mobile platforms

Zhu

Liang

2023

AEAT

View full text Add to dashboard Cite

show abstract

“…Besides, infrared and sonar sensors detect object only within one meter range [2]. Vision based sensor consists of stereo camera -based obstacle detection [19] [20] [21] and monocular camerabased obstacle detection sensor [22][23] [24] [25]. Camera sensor is a passive sensor [26] that have poor object distance detection but, contain huge amount of information that can be further analyze in terms features of edges [27] [28], point [29] and grayscale values [30] to detect object.…”

Section: Introductionmentioning

confidence: 99%

Region Detection Technique Using Image Subtraction and Pixel Expansion Cue for Obstacle Detection System on Small – Sized UAV

Wafi Abdul Aziz,

Faiz Ramli

2024

IJCDS

View full text Add to dashboard Cite

This research paper is about method of detection of free region and obstacle region by combining image segmentation and frame subtraction method. The application will be further study to be used by Unmanned Aerial Vehicles (UAVs). This method intends to minimize the weight of UAV by avoiding heavy sensors. Objective of this research is to utilize the pixel expansion of object to find free region. K-means segmentation will be used to separate the interest area from the background. Then, segmented image frame will be subtracted and then divided into several grids and the amount of subtracted pixel that has been set into yellow and black color of each grid will be calculated with respect to distance given. The expansion of pixel will be detected as, the distance between image frame coming closer, number of obstacle pixel will be higher. The application of simple LIDAR that emits single ray to frontal obstacle will initiate the camera to capture image frame to be further analyze. Experiment was carried out in close environment with different cases and total of 100 images has been captured that consist of texture obstacle, texture-less obstacle, and multiple obstacles. The findings showed bearable results as the free region detection is 88.0% for texture obstacle and up to 84.0% of free region were successfully detected for texture-less obstacle. Due to lack of cues and texture in texture-less object, the algorithm had difficulties detecting the center of object that expands in static form.

show abstract

“…In recent years, with the continuous advancement of deep learning, neural networks have emerged as the predominant approach in the field of object detection [6]. The application in the aerospace domain holds the potential to expedite the intelligent evolution of space situational awareness [7][8][9]. However, the formidable cost and complexity associated with space data collection have led to the generation of most space target datasets through model simulations, overlooking the intricacies of real-space imaging [10].…”

Section: Introductionmentioning

confidence: 99%

TYCOS: A Specialized Dataset for Typical Components of Satellites

Bian,

Cao,

Zhang

et al. 2024

Applied Sciences

View full text Add to dashboard Cite

The successful detection of key components within satellites is a crucial prerequisite for executing on-orbit capture missions. Due to the inherent data-driven functionality, deep learning-based component detection algorithms rely heavily on the scale and quality of the dataset for their accuracy and robustness. Nevertheless, existing satellite image datasets exhibit several deficiencies, such as the lack of satellite motion states, extreme illuminations, or occlusion of critical components, which severely hinder the performance of detection algorithms. In this work, we bridge the gap via the release of a novel dataset tailored for the detection of key components of satellites. Unlike the conventional datasets composed of synthetic images, the proposed Typical Components of Satellites (TYCOS) dataset comprises authentic photos captured in a simulated space environment. It encompasses three types of satellite, three types of key components, three types of illumination, and three types of motion state. Meanwhile, scenarios with occlusion in front of the satellite are also taken into consideration. On the basis of TYCOS, several state-of-the-art detection methods are employed in rigorous experiments followed by a comprehensive analysis, which further enhances the development of space scene perception and satellite safety.

show abstract

Dataset generation and validation for spacecraft pose estimation via monocular images processing

Cited by 13 publications

References 17 publications

A satellite component contour extraction method for lightweight space mobile platforms

A satellite component contour extraction method for lightweight space mobile platforms

Region Detection Technique Using Image Subtraction and Pixel Expansion Cue for Obstacle Detection System on Small – Sized UAV

TYCOS: A Specialized Dataset for Typical Components of Satellites

Contact Info

Product

Resources

About