Image Tone Control via Weighted Object Region Luminance

Resolve Bottlenecks,
Find Innovative Solutions
Generate Solutions

Solution Overview

Problem

Conventional image processing techniques fail to effectively suppress blown-out highlights and blocked-up shadows in regions other than face areas in captured images, such as flowers, animals, or still subjects, due to insufficient dynamic range of image sensors.

Innovation Solution

An image processing apparatus and method that detects multiple object regions, identifies a main object region, calculates representative luminance values and distributions, determines tone characteristics, and applies tone corrections to improve the main object region's tones without impairing other regions, using a weighted approach to control luminance and histogram adjustments.

Engineering Contradictions & Design Principles

VSEngineering Contradiction Analysis

1Manufacturing precision

If exposure conditions are optimized for the main object region, then tone quality of the main object is improved, but blown-out highlights or blocked-up shadows occur in other regions with different luminance characteristics

Engineering Contradiction:
Improvetone qualityVSAvoidblown-out highlights and blocked-up shadows
Core Design Contradiction:
Manufacturing precisionVSObject-affected harmful factors

Solution Approach 1:

The image is divided into multiple object regions with different luminance characteristics. The tone correction is applied separately to each region based on its specific luminance distribution, allowing the main object region to be optimized without causing blown-out highlights or blocked-up shadows in other regions.

Inventive Principle:
Principle #1Segmentation

Solution Approach 2:

Different tone correction characteristics are applied to different object regions based on their local luminance properties. The main object region receives tone correction optimized for its luminance distribution, while other regions receive appropriate tone correction based on their own luminance characteristics, preventing harmful effects.

Inventive Principle:
Principle #3Local quality

2Manufacturing precision

If conventional face-specific tone control is applied, then tone quality of face regions is improved, but objects other than faces (e.g., flowers, animals) cannot have their blown-out highlights or blocked-up shadows suppressed

Engineering Contradiction:
Improvetone qualityVSAvoidapplicability to different object types
Core Design Contradiction:
Manufacturing precisionVSAdaptability or versatility

Solution Approach 1:

The tone correction apparatus is designed to handle multiple types of objects universally. Instead of being limited to face-specific control, the system can detect and apply appropriate tone correction to any object region including flowers, animals, and other subjects by analyzing their luminance distributions and applying region-specific correction characteristics.

Inventive Principle:
Principle #6Universality (Multi-functionality)

Data Source

PatentUS10108878B2Image processing apparatus, image processing method, and storage medium for tone control of each object region in an image
Publication Date: 2018.10.23 CANON KK
  • US10108878B2 patent drawing
  • US10108878B2 patent drawing
  • US10108878B2 patent drawing

AI summary

An image processing apparatus includes: an object region detection unit; a main object region detection unit; a representative luminance value calculation unit configured to calculate one representative luminance value of each object region, by weighting the main object region in that object region; a luminance distribution calculation unit configured to calculate a luminance distribution in each object region, by weighting the main object region in that object region; a tone characteristics determination unit configured to determine tone characteristics for controlling tones of the input image, based on the representative luminance values of the object regions and the luminance distributions in the object regions; and a tone correction unit configured to correct the tones of the input image, using the tone characteristics determined by the tone characteristics determination unit.