Image Processing Method for Overexposed Facial Regions

Solution Overview

Problem

Current image processing technologies face challenges in accurately enhancing light effects on images with overexposed facial regions, leading to potential distortion and reduced processing accuracy.

Innovation Solution

An image processing method that detects overexposed regions in facial images, calculates a light effect intensity coefficient, and applies a target light effect model to perform light enhancement processing, adjusting the intensity based on the overexposed region to prevent distortion and improve accuracy.

Engineering Contradictions & Design Principles

VSEngineering Contradiction Analysis

1Illumination intensity

If light enhancement processing is applied to images with overexposed facial regions, then the light effect enhancement is improved, but the distortion in the face region increases

Engineering Contradiction:

Improvelight effect enhancementVSAvoidface region distortion

Core Design Contradiction:

Illumination intensityVSManufacturing precision

Solution Approach 1:

The patent applies different light effect models to different regions of the image. Specifically, it detects overexposed regions in facial areas and applies a first light effect model to non-overexposed regions while applying a second light effect model to overexposed regions. This local differentiation allows the system to enhance light effects where appropriate while avoiding distortion in overexposed facial regions.

Inventive Principle:

Principle #3Local quality

Solution Approach 2:

The patent dynamically selects between different light effect models based on the detected overexposed regions. The system adjusts the processing approach in real-time by choosing between a first light effect model (for non-overexposed regions) and a second light effect model (for overexposed regions), thereby adapting the enhancement intensity and type to the specific characteristics of each region.

Inventive Principle:

Principle #15Dynamics

2Illumination intensity

If light enhancement processing is applied to images with overexposed facial regions, then the light effect enhancement is improved, but the processing accuracy decreases

Engineering Contradiction:

Improvelight effect enhancementVSAvoidimage processing accuracy

Core Design Contradiction:

Illumination intensityVSMeasurement precision

Solution Approach 1:

The patent applies different light effect models to different regions of the image. Specifically, it detects overexposed regions in facial areas and applies a first light effect model to non-overexposed regions while applying a second light effect model to overexposed regions. This local differentiation allows the system to enhance light effects where appropriate while avoiding distortion in overexposed facial regions.

Inventive Principle:

Principle #3Local quality

Solution Approach 2:

The patent incorporates feedback through the detection of overexposed regions, which guides the selection of appropriate light effect models. By continuously monitoring the image for overexposed areas and adjusting the processing parameters accordingly, the system improves overall processing accuracy while maintaining effective light enhancement.

Inventive Principle:

Principle #23Feedback

Data Source

PatentEP3849170B1Image processing method, electronic device, and computer-readable storage medium

Publication Date: 2023.12.20 GUANGDONG OPPO MOBILE TELECOMMUNICATIONS CORP LTD

Figure 1
Figure 2
Figure 3

AI summary

An image processing method, comprising: obtaining an image to be processed; detecting a face region in said image and detecting an overexposure region in the face region; obtaining a luminous efficacy intensity coefficient according to the overexposure region, and obtaining a target luminous efficacy model according to the luminous efficacy intensity coefficient, the target luminous efficacy model being a model for simulating the change in the light; and performing, according to the luminous efficacy model, luminous efficacy enhancement on said image.