Moving object detection from multi-depth images with an attention-enhanced CNN

📅 2025-12-04
📈 Citations: 0
Influential: 0
📄 PDF
🤖 AI Summary
Manual verification in wide-field solar system surveys severely limits the efficiency of moving object detection. Method: This paper proposes an end-to-end multi-input convolutional neural network (CNN) architecture incorporating the Convolutional Block Attention Module (CBAM), enabling direct processing of multi-depth, multi-frame stacked images and adaptive enhancement of salient motion features in both spatial and channel dimensions. The method eliminates traditional manual screening, achieving fully automated detection and classification from raw image sequences. Results: Evaluated on approximately 2,000 real survey images, the model achieves 98.9% accuracy and an AUC of 0.992. With optimized detection thresholds, it maintains high recall while reducing human verification effort by over 99%. This work significantly advances the automation level and operational efficiency of moving object discovery in astronomical surveys.

Technology Category

Application Category

📝 Abstract
One of the greatest challenges for detecting moving objects in the solar system from wide-field survey data is determining whether a signal indicates a true object or is due to some other source, like noise. Object verification has relied heavily on human eyes, which usually results in significant labor costs. In order to address this limitation and reduce the reliance on manual intervention, we propose a multi-input convolutional neural network integrated with a convolutional block attention module. This method is specifically tailored to enhance the moving object detection system that we have developed and used previously. The current method introduces two innovations. This first one is a multi-input architecture that processes multiple stacked images simultaneously. The second is the incorporation of the convolutional block attention module which enables the model to focus on essential features in both spatial and channel dimensions. These advancements facilitate efficient learning from multiple inputs, leading to more robust detection of moving objects. The performance of the model is evaluated on a dataset consisting of approximately 2,000 observational images. We achieved an accuracy of nearly 99% with AUC (an Area Under the Curve) of >0.99. These metrics indicate that the proposed model achieves excellent classification performance. By adjusting the threshold for object detection, the new model reduces the human workload by more than 99% compared to manual verification.
Problem

Research questions and friction points this paper is trying to address.

Detects moving objects in solar system survey data
Reduces reliance on manual verification by human eyes
Enhances detection accuracy using attention-based CNN architecture
Innovation

Methods, ideas, or system contributions that make the work stand out.

Multi-input CNN processes multiple stacked images simultaneously
Convolutional block attention module focuses on spatial and channel features
Model reduces human workload by over 99% with near 99% accuracy
🔎 Similar Papers
No similar papers found.
💼 Related Jobs
No related jobs found.
M
Masato Shibukawa
The Graduate University for Advanced Studies, SOKENDAI, Shyonan International Village, Hayama, Miura, Kanagawa 240-0193, Japan
F
Fumi Yoshida
University of Occupational and Environmental Health, Japan, 1-1 Iseigaoka, Yahata, Kitakyusyu, Fukuoka 807-8555, Japan
T
Toshifumi Yanagisawa
Star Signal Solutions Inc, Jindaiji Higashimachi 7-44-1, Chofu, Tokyo 182-8522, Japan
T
Takashi Ito
Planetary Exploration Research Center, Chiba Institute of Technology, 2-17-1 Tsudanuma, Narashino, Chiba 275-0016, Japan
H
Hirohisa Kurosaki
Chofu Headquarters, Japan Aerospace Exploration Agency, Jindaiji Higashimachi 7-44-1, Chofu, Tokyo 182-0012, Japan
M
Makoto Yoshikawa
ISAS, Japan Aerospace Exploration Agency, Yoshinodai 3-1-1, Sagamihara, Kanagawa 252-0222, Japan
K
Kohki Kamiya
Chofu Headquarters, Japan Aerospace Exploration Agency, Jindaiji Higashimachi 7-44-1, Chofu, Tokyo 182-0012, Japan
J
Ji-an Jiang
Department of Astronomy, University of Science and Technology of China, Hefei 230026, China
W
Wesley Fraser
National Research Council of Canada, Herzberg Astronomy and Astrophysics Research Centre, 5071 W. Saanich Rd., Victoria, BC, V9E 2E7, Canada
J
JJ Kavelaars
Department of Physics and Astronomy, University of Victoria, Elliott Building, 3800 Finnerty Road, Victoria, BC, V8P 5C2, Canada
S
Susan Benecchi
Planetary Science Institute, 1700 East Fort Lowell, Suite 106, Tucson, AZ 85719, USA
A
Anne Verbiscer
Southwest Research Institute, 1050 Walnut Street, Boulder, CO 80302, USA
A
Akira Hatakeyama
The Graduate University for Advanced Studies, SOKENDAI, Shyonan International Village, Hayama, Miura, Kanagawa 240-0193, Japan
H
Hosei O
University of Tokyo, 7-3-1 Hongo, Bunkyo, Tokyo 113-8654, Japan
N
Naoya Ozaki
The Graduate University for Advanced Studies, SOKENDAI, Shyonan International Village, Hayama, Miura, Kanagawa 240-0193, Japan