Background Object Feature Detection for Video Analysis
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Existing object detection methods, such as background subtraction, often incorrectly detect changes in background objects as new objects due to movement or rotation, leading to frequent detection errors.
Innovation Solution
An object detection apparatus and method that includes a video input unit, an object region detection unit, a selection unit, a generation unit, and a determination unit to identify and generate background object feature information, allowing for the differentiation between background objects and new objects based on pre-existing features.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Reliability
If background subtraction method is used to detect objects, then object detection can be performed, but detection errors occur when background objects move or change
Solution Approach 1:
The patent segments the detection process into multiple stages: initial object region detection, background object candidate identification, feature amount extraction, and determination. By dividing the complex detection task into manageable segments, the system improves reliability without overwhelming complexity at any single stage.
Solution Approach 2:
The patent performs preliminary actions by pre-storing feature amounts of background objects in a storage unit before actual detection occurs. This preliminary preparation enables faster and more accurate determination during operation, improving detection reliability while maintaining system efficiency.
2Reliability
If feature amounts are compared to distinguish background objects from new objects, then detection errors are reduced, but processing time increases
Solution Approach 1:
The patent extracts only the necessary feature amounts (color, shape, texture) from background object candidate regions and stores them for comparison. By extracting only relevant features rather than processing entire images, the system maintains high detection accuracy while reducing processing time.
Solution Approach 2:
The patent applies local quality by focusing feature extraction and comparison only on candidate background object regions rather than the entire image. This localized approach maintains detection accuracy for background objects while significantly reducing overall processing time.
3Adaptability or versatility
If new features appear upon movement or rotation of background objects, then the object changes appearance, but existing detection methods fail to recognize it as a background object
Solution Approach 1:
The patent implements dynamics by continuously updating the stored feature amounts of background objects based on new observations. When a background object moves or rotates and presents new features, the system updates its stored representation, enabling it to adapt to changes while maintaining reliable detection accuracy.
Solution Approach 2:
The patent employs feedback mechanisms where detection results are used to update the background object feature database. When new features are detected from moving background objects, this information feeds back into the system to refine future comparisons, improving both adaptability and reliability over time.
Data Source
AI summary
The object detection apparatus prevents or eliminates detection errors caused by changes of an object which frequently appears in a background. To this end, an object detection apparatus includes a detection unit which detects an object region by comparing an input video from a video input device and a background model, a selection unit which selects a region of a background object originally included in a video, a generation unit which generates background object feature information based on features included in the background object region, and a determination unit which determines whether or not the object region detected from the input video is a background object using the background object feature information.