Moving Object Detection via Trajectory Stationary Measures
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Conventional methods for detecting moving objects, especially pedestrians with deformation, struggle when multiple persons are walking and there are significant changes in posture and size, leading to inaccurate object detection due to the need for numerous parameters and issues with background changes when images are captured by a moving camera.
Innovation Solution
A moving object detection apparatus that calculates a stationary measure for each trajectory, determining the likelihood of it belonging to a stationary object, and uses this measure to transform and separate trajectories of moving objects from stationary objects based on distance ratios, ensuring accurate detection even with changing shapes and camera motion.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Adaptability or versatility
If conventional methods use multiple parameters to detect moving objects with deformation, then detection coverage is improved, but detection accuracy deteriorates due to parameter complexity and background changes
Solution Approach 1:
The patent segments the detection process into trajectory extraction, stationary measure calculation, and distance-based classification. By dividing the complex detection task into distinct stages, each handling specific aspects (motion patterns, stationary likelihood, spatial relationships), the system achieves both comprehensive coverage and high accuracy without being overwhelmed by parameter complexity
Solution Approach 2:
The patent introduces a stationary measure as an intermediary metric that quantifies the likelihood of trajectories belonging to stationary objects. This intermediary measure acts as a bridge between raw trajectory data and final classification, enabling accurate distinction between moving and stationary objects even in complex scenes with camera motion and deformable targets
2Productivity
If the camera is moving, then the ability to capture dynamic scenes is improved, but the difficulty of detecting and measuring moving objects increases due to background changes
Solution Approach 1:
The patent employs dynamic trajectory analysis that adapts to camera motion by tracking object paths across multiple frames and calculating stationary measures based on motion patterns rather than absolute positions. This dynamic approach allows the system to maintain detection accuracy regardless of camera movement, as it focuses on relative motion characteristics rather than fixed spatial references
Solution Approach 2:
The patent performs preliminary trajectory extraction and stationary measure calculation for all detected paths before final classification. By pre-processing the data to identify motion patterns and calculate stationary likelihoods in advance, the system reduces the computational complexity of real-time detection and improves accuracy even when the camera is moving through dynamic scenes
Data Source
AI summary
A moving object detection apparatus includes: a stationary measure calculation unit calculating, for each of trajectories, a stationary measure representing likelihood that the trajectory belongs to a stationary object; a distance calculation unit calculating a distance representing similarity between trajectories; and a region detection unit (i) performing a transformation based on the stationary measures and the distances between the trajectories, so that a ratio of a distance between a trajectory on stationary object and a trajectory on moving object, to a distance between trajectories both belonging to stationary object becomes greater than a ratio obtained before the transformation and (ii) detecting the moving object region by separating the trajectory on the moving object from the trajectory on the stationary object, based on a geodesic distance between the trajectories.