Space Debris Detection via Image Grouping and Median Evaluation
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Current methods for detecting moving objects from captured images, such as space debris, require extensive processing due to the need to evaluate median pixel values across multiple images, leading to time-consuming calculations and high probabilities of speculative execution failures, especially when dealing with noise components like stars or cosmic rays.
Innovation Solution
A method that divides images into groups and evaluates pixel values using robust statistics and average values, reducing the number of operation processes by estimating movement and calculating evaluation values within each group, thereby detecting moving objects with less computational effort and lower failure probabilities.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Measurement precision
If the stacking method is used to detect moving objects by evaluating median pixel values across multiple images, then the detection accuracy is improved and noise components are suppressed, but the computational complexity and processing time increase significantly
Solution Approach 1:
The patent divides the stack of multiple images into multiple groups, where each group contains a subset of the total images. The detection process is performed separately for each group, evaluating median pixel values within each group rather than across all images simultaneously. This segmentation reduces the computational complexity from O(n log n) for the entire stack to O(k log m) for each group, where n is the total number of images, k is the number of groups, and m is the number of images per group.
Solution Approach 2:
The patent performs the stacking operation partially by processing only a subset of images in each group rather than requiring all images to be processed together. This partial action allows the detection process to proceed with reduced computational burden while still achieving effective noise suppression and moving object detection through the median value evaluation within each group.
2Reliability
If the stacking method processes all images simultaneously to evaluate median pixel values, then comprehensive noise suppression is achieved, but the processing time becomes excessively long
Solution Approach 1:
The patent segments the image stack into multiple groups and processes each group independently and in parallel. This segmentation enables the noise suppression function to be distributed across multiple processing units, reducing the overall processing time while maintaining the effectiveness of median value-based noise suppression within each group.
Solution Approach 2:
The patent applies periodic action by processing groups of images in sequential batches rather than all at once. Each group is processed in a periodic manner, allowing the system to complete detection tasks in multiple shorter time intervals rather than one excessively long processing period, thereby reducing the perceived processing time while maintaining comprehensive noise suppression.
3Measurement precision
If candidate areas are evaluated by trial and error with various cropping positions and staggering amounts, then accurate superposition of moving objects is achieved, but the processing work becomes enormous
Solution Approach 1:
The patent performs preliminary action by pre-dividing the image stack into groups with predetermined cropping areas and staggering amounts. Instead of evaluating all possible cropping positions and staggering amounts through trial and error, the method preliminarily establishes group structures that facilitate accurate superposition, thereby reducing the processing work required while maintaining high superposition accuracy for moving object detection.
Data Source
AI summary
A moving object detection method according to an embodiment performs a stacking process step of: dividing captured images into p groups each including m successive images; and for each group, applying a stacking method to the m images. Subsequently, for each group, the method extracts a candidate for space debris based on an evaluation value for space debris detection, which is obtained from pixel values at the same pixel positions overlapping one another across the m images superposed one another. Thereafter, the method judges whether the extracted candidate for the space debris in each group is chosen as a final candidate for the space debris based on the coincidence or proximity in terms of the movement direction and the moving velocity.