Moving Object Detection via Corner Feature Point Clustering
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Conventional moving object detection methods using image alignment and visual optical flow tracing techniques face challenges such as high costs and reduced image processing speed due to the need for multiple cameras and contour-based approaches, which are less accurate and more prone to noise compared to region-based methods.
Innovation Solution
A moving object detection system comprising an image capturing device and a controlling and processing module with a pre-processor, feature extraction module, image optical flow estimation module, feature points grouping module, and moving object determination module, utilizing patch-based feature difference calculations, structure tensor matrices, and Lucas-Kanade optical flow to enhance computing speed and accuracy.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Measurement precision
If visual optical flow tracing technique with multiple cameras is used, then measurement precision of moving object detection is improved, but device complexity and cost increase
Solution Approach 1:
The patent segments the detection task by dividing the image into multiple blocks and processing each block independently to calculate optical flow. This allows a single camera to achieve detection accuracy comparable to multi-camera systems by focusing computational effort on specific regions of interest rather than requiring multiple simultaneous camera views.
Solution Approach 2:
The patent introduces an intermediary processing stage that calculates optical flow vectors for each block and uses these vectors to identify moving objects. This intermediary optical flow calculation acts as a mediator that transforms single-camera sequential images into motion information that would traditionally require multiple cameras to capture simultaneously.
2Device complexity
If contour-based image alignment approach is used, then device complexity is reduced, but measurement precision and noise resistance deteriorate
Solution Approach 1:
The patent applies local quality by treating different blocks of the image differently - each block is processed independently with its own optical flow calculation and motion detection. This local processing approach improves precision in each region while maintaining overall system simplicity, overcoming the limitations of global contour-based alignment.
3Measurement precision
If patch-based feature difference calculation with structure tensor is used, then measurement precision is improved, but computing time increases
Solution Approach 1:
The patent divides the image into multiple blocks and processes each block independently for feature detection. This segmentation allows parallel processing of different regions, reducing overall computing time while maintaining the precision benefits of patch-based structure tensor calculations in each local area.
Solution Approach 2:
The patent applies partial action by focusing computational resources only on blocks that contain potential moving objects, rather than processing the entire image uniformly. This selective processing reduces total computation time while maintaining detection precision in critical areas.
Data Source
Figure 1
Figure 2
Figure 3
AI summary
Disclosures of the present invention describe a moving object detection system and method, wherein a pre-processer module, a feature extraction module, an image optical flow estimation module, a feature points grouping module, and a moving object determination module are provided in a controlling and processing module of the system by a form of library, variables, or operands. Moreover, a feature difference calculation unit, a matrix establishing unit and a corner feature point acquiring unit are provided in the feature extraction module, and that is helpful for enhancing computing speed of the controlling and processing device in verifying corner feature points from image frames. Therefore, after the corner feature points are applied with a cluster labeling process, the moving object determination module can achieve motion detection of at least one object locating in a monitoring area by determining whether corner feature point groups move or not.