Camera Exposure Estimation Using Neural Networks to Avoid Exposure Hunting

Resolve Bottlenecks,
Find Innovative SolutionsGenerate Solutions

Solution Overview

Problem

Existing camera technologies face issues with exposure hunting, leading to inefficient exposure parameter calculation, prolonged image processor usage, and reduced battery life due to frequent exposure stabilization, especially when used as a light sensor alternative.

Innovation Solution

A method utilizing a pre-trained neural network to determine optimal exposure parameters and scene brightness using a single image frame, reducing exposure setting time and hardware utilization.

Engineering Contradictions & Design Principles

VSEngineering Contradiction Analysis

1Measurement precision

If conventional exposure calculation methods are used, then accurate exposure parameters are obtained, but the time required for achieving optimal exposure is approximately 250ms-300ms

Engineering Contradiction:

Improveexposure parameter accuracyVSAvoidexposure stabilization time

Core Design Contradiction:

Measurement precisionVSLoss of time

Solution Approach 1:

The patent pre-calculates and stores exposure parameters in a lookup table based on histogram data before actual image capture. When a scene is captured, the system quickly queries the pre-computed table rather than performing real-time exposure calculations, thereby achieving accurate exposure parameters without the 250-300ms stabilization delay.

Inventive Principle:

Principle #10Preliminary action

2Stability of the object's composition

If exposure hunting is avoided using conventional methods, then exposure stability is improved, but the image processor runs longer and consumes more computational power

Engineering Contradiction:

Improveexposure stabilityVSAvoidimage processor energy consumption

Core Design Contradiction:

Stability of the object's compositionVSUse of energy by moving object

Solution Approach 1:

The patent replaces the iterative computational exposure hunting process with a direct table lookup mechanism. Instead of having the image processor repeatedly calculate and adjust exposure parameters, the system substitutes this computational mechanical process with a simple data retrieval operation from pre-computed tables, significantly reducing processor energy consumption while maintaining exposure stability.

Inventive Principle:

Principle #28Mechanics substitution (Replace mechanical system)

3Adaptability or versatility

If the camera is used as a light sensor alternative with frequent operations, then scene brightness measurement capability is improved, but the camera lifespan and battery life are reduced due to over-utilization

Engineering Contradiction:

Improvelight sensor functionalityVSAvoidcamera lifespan

Core Design Contradiction:

Adaptability or versatilityVSDuration of action of stationary object

Solution Approach 1:

The patent pre-computes and stores exposure parameters in lookup tables before the camera is actually used for light sensing operations. This preliminary preparation allows the camera to perform rapid table-lookup-based exposure determination during frequent light sensing operations, minimizing processing time and reducing wear on the camera hardware and battery, thereby extending their lifespan while maintaining versatility.

Inventive Principle:

Principle #10Preliminary action

Data Source

PatentEP4011060B1Method of fast estimation of scene brightness and optimal camera exposure

Publication Date: 2026.03.18 SAMSUNG ELECTRONICS CO LTD

Figure 1A~1B
Figure 2A
Figure 2B

AI summary

A method and an apparatus for adjusting an optimal exposure of a camera are provided. The method may include: receiving a camera start command to activate a camera of an electronic device; capturing an image frame of a scene by the camera; obtaining an Exposure Brightness Response (EBR) of the camera; providing the EBR and the image frame as an input to a pre-trained neural network specific to the camera; and determining a plurality of exposure parameters of the camera and a mean pixel brightness (MPB) by using the pre-trained neural network.