Composition-Based Exposure Measurement for Image Correction
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Current image correction methods inaccurately measure the exposure degree of objects in images due to measuring across the entire image, leading to inefficient CPU usage and incorrect object detection, as seen in methods that detect high-frequency components, which can mistakenly identify background as the object.
Innovation Solution
A composition-based exposure measuring method and apparatus that uses a computer processor to measure the exposure degree of an object by determining a region based on image composition, such as elliptical or golden-section compositions, and calculates exposure amounts using Gaussian distribution weighting or histogram analysis to accurately determine the object's exposure degree.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Measurement precision
If exposure degree is measured across the entire image region, then measurement coverage is complete, but measurement precision deteriorates due to inclusion of background regions
Solution Approach 1:
The patent divides the image into multiple regions of interest (ROIs) based on composition rules (e.g., rule of thirds, golden ratio) rather than measuring the entire image. This segmentation isolates the main object from the background, enabling precise exposure measurement of only the relevant areas while excluding irrelevant background regions from the measurement calculation.
Solution Approach 2:
The patent applies different measurement strategies to different regions of the image based on their importance. By identifying and prioritizing specific composition-based regions where the main object is likely to be located, the system concentrates measurement resources on areas that matter most for accurate exposure assessment, rather than treating all pixels equally.
2Measurement precision
If object detection methods are used to identify the main object, then exposure measurement accuracy improves, but CPU cost increases significantly
Solution Approach 1:
The patent pre-defines regions of interest based on composition rules before performing any exposure measurement. By establishing these composition-based ROIs in advance using simple geometric calculations rather than complex object detection algorithms, the system prepares the measurement framework with minimal computational overhead, avoiding the need for expensive real-time object detection while still focusing on the correct areas.
Solution Approach 2:
The patent uses simple, computationally inexpensive composition-based region definitions instead of complex, resource-intensive object detection algorithms. These composition rules provide sufficient accuracy for exposure measurement without requiring the heavy computational resources of advanced object recognition systems, effectively replacing expensive methods with cheaper alternatives that meet the functional requirements.
3Speed
If high-frequency component detection is used to identify objects, then detection speed improves, but reliability deteriorates due to false detection of background regions
Solution Approach 1:
The patent introduces composition rules as an intermediary framework that guides exposure measurement without requiring direct object detection. Instead of using high-frequency component detection that leads to false positives, the system uses composition-based region definitions as a mediator to identify where the main object is likely to be located, combining the speed of rule-based approaches with the reliability of focused measurement.
Data Source
AI summary
Provided is a composition-based exposure measuring method and apparatus for measuring an exposure degree of an object included in an image, including: receiving an input of the image; measuring an exposure amount of a pixel located in a region determined based on a composition, among pixels of the received image; and determining the exposure degree of the object based on the measured exposure amount.