Automatic ISP Tuning Using Reference Images

Resolve Bottlenecks,
Find Innovative SolutionsGenerate Solutions

Solution Overview

Problem

Conventional techniques for tuning image signal processors (ISPs) are labor-intensive and time-consuming, requiring manual effort from IQ engineers, which can take several weeks to deliver imaging solutions, and are prone to human errors and inefficiencies, especially for low-tiered customers.

Innovation Solution

An automatic image quality tuning mechanism using nonlinear optimization and reference images is employed, which generates reference images for each ISP functionality block, allowing for the automatic tuning of ISPs through parameter abstraction, fitness calculation, and hybrid optimization, reducing the effort and time needed to achieve high-quality images.

Engineering Contradictions & Design Principles

VSEngineering Contradiction Analysis

1Manufacturing precision

If manual tuning by IQ engineers is used, then image quality can be optimized, but the process becomes labor-intensive and time-consuming

Engineering Contradiction:

Improveimage qualityVSAvoidtuning time

Core Design Contradiction:

Manufacturing precisionVSLoss of time

Solution Approach 1:

The system performs automatic ISP tuning by having the computer automatically generate reference images, calculate fitness metrics, and optimize parameters without requiring manual intervention from IQ engineers. The automated tuning mechanism enables the system to self-adjust and self-optimize, eliminating the need for labor-intensive manual tuning processes while maintaining high image quality standards.

Inventive Principle:

Principle #25Self-service

Solution Approach 2:

The patent replaces the manual mechanical tuning process with an automated computational system. Instead of engineers manually adjusting parameters, the system uses computer-based algorithms to automatically generate reference images, compute fitness calculations, and optimize ISP parameters, substituting human manual operations with automated mechanical and computational processes.

Inventive Principle:

Principle #28Mechanics substitution (Replace mechanical system)

2Manufacturing precision

If manual tuning by IQ engineers is used, then image quality can be optimized, but the process requires great deal of efforts

Engineering Contradiction:

Improveimage qualityVSAvoidtuning effort

Core Design Contradiction:

Manufacturing precisionVSEase of operation

Solution Approach 1:

Inventive Principle:

Principle #25Self-service

Solution Approach 2:

Inventive Principle:

Principle #28Mechanics substitution (Replace mechanical system)

3Productivity

If conventional manual tuning is used, then imaging solutions can be delivered, but human errors and inefficiencies occur

Engineering Contradiction:

Improvesolution deliveryVSAvoiderror rate

Core Design Contradiction:

ProductivityVSReliability

Solution Approach 1:

Inventive Principle:

Principle #25Self-service

Solution Approach 2:

The system implements a feedback mechanism where fitness calculations are performed by comparing processed images against reference images, and the results are used to automatically adjust and optimize ISP parameters. This closed-loop feedback system ensures continuous improvement and eliminates human error by using objective computational metrics rather than subjective manual judgment.

Inventive Principle:

Principle #23Feedback

4Loss of time

If automatic tuning is implemented, then time and effort are reduced, but complex optimization processes are required

Engineering Contradiction:

Improvetuning timeVSAvoidoptimization complexity

Core Design Contradiction:

Loss of timeVSDevice complexity

Solution Approach 1:

The automatic tuning process is segmented into distinct functional modules: reference image generation, fitness calculation, and parameter optimization. Each module handles a specific aspect of the tuning process, making the overall complex system manageable through modular decomposition. This segmentation allows the system to process multiple ISP functionality blocks independently and systematically.

Inventive Principle:

Principle #1Segmentation

Solution Approach 2:

The automatic tuning mechanism is designed to be universal and applicable to multiple ISP functionality blocks simultaneously. The same automated framework handles different image processing functions (such as demosaicing, noise reduction, sharpening) using a unified approach, reducing the need for separate tuning processes for each function and simplifying the overall system architecture.

Inventive Principle:

Principle #6Universality (Multi-functionality)

Data Source

PatentUS10755425B2Automatic tuning of image signal processors using reference images in image processing environments

Publication Date: 2020.08.25 INTEL CORP

AI summary

A mechanism is described for facilitating automatic tuning of image signal processors using reference images in image processing environments, according to one embodiment. A method of embodiments, as described herein, includes one or more processors to: receive images associated with one or more scenes captured by one or more cameras; access tuning parameters associated with functionalities within an image signal processor (ISP) pipeline; generate reference images based on the tuning parameters, wherein a reference image is associated with an image for each functionality within the ISP pipeline; and automatically tune the ISP pipeline based on selection of one or more of the reference images for one or more of the images for one or more of the functionalities.