Imaging Apparatus Subject Tracking via Feature Value Extraction
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Existing imaging apparatuses using visible light communication struggle to accurately track subjects with light-emitting units, as they often rely on maintaining the light source at the image center, which can be influenced by luminance variations and shape, leading to instability in tracking.
Innovation Solution
The apparatus includes an imaging unit, an area detecting unit, a position detecting unit, a local area searching unit, and a tracking unit that identifies and follows a local area with a high feature value within a prescribed region, excluding the influence of light blinking, allowing for stable and accurate subject tracking.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Ease of operation
If the light emitting unit is maintained at the center of the image for tracking, then the tracking process is simplified, but tracking accuracy deteriorates due to luminance variations and light shape changes
Solution Approach 1:
The invention extracts the feature amount information from the light emitting unit's image data, separating the tracking target from the problematic luminance variations and light shape changes. By using extracted feature amounts rather than raw pixel data, the system achieves accurate tracking without being influenced by lighting conditions.
Solution Approach 2:
The invention changes the parameter used for tracking from luminance-based pixel values to feature amount-based descriptors. This parameter transformation allows the system to track the light emitting unit's position accurately while being invariant to luminance variations and changes in light shape.
2Speed
If the light emitting unit's position is tracked directly, then tracking speed is maintained, but tracking stability deteriorates due to influence from light blinking and shape variations
Solution Approach 1:
The invention implements feedback by continuously calculating feature amounts from current image data and comparing them with reference feature amounts. This feedback mechanism allows the system to maintain accurate and stable tracking by adjusting based on the difference between current and reference feature amounts, rather than relying on simple position maintenance.
Solution Approach 2:
The invention replaces the mechanical approach of maintaining the light source at the image center with an algorithmic feature-based tracking system. This substitution uses feature amount calculation and comparison to determine tracking adjustments, providing both speed and stability through computational methods rather than mechanical positioning.
Data Source
AI summary
An imaging apparatus for tracking a subject is provided. The An imaging apparatus has an imaging unit which images a subject to obtain an image, an area detecting unit which detects in the image obtained by the imaging unit a light changing area, in which light changes in a time-series manner, a position detecting unit which detects a position of the light changing area detected by the area detecting unit in the image obtained by the imaging unit, a local area searching unit which searches for a local area within a prescribed region defined on the basis of the position of the light changing area detected by the position detecting unit, wherein the local area shows a large feature value in an image processing, and a tracking unit which tracks movement of the local area found by the local area searching unit.