Camera Phase Detection Depth Mapping with Lens Calibration

Resolve Bottlenecks,
Find Innovative SolutionsGenerate Solutions

Solution Overview

Problem

Existing camera systems struggle to accurately generate depth maps using phase detect sensors due to limitations in correlating phase difference measurements with lens defocus and object distances, which affects autofocus performance.

Innovation Solution

A method involving a phase detect sensor that utilizes pre-stored calibration data to correlate phase difference measurements with lens defocus and object distances, enabling the generation of a depth map by determining appropriate lens positions and object distances based on these measurements.

Engineering Contradictions & Design Principles

VSEngineering Contradiction Analysis

1Measurement precision

If phase difference measurements are used for autofocus, then autofocus capability is provided, but accurate depth map generation is difficult due to limited correlation with lens defocus and object distances

Engineering Contradiction:

Improvedepth map accuracyVSAvoidcalibration data structure

Core Design Contradiction:

Measurement precisionVSDevice complexity

Solution Approach 1:

The patent performs preliminary calibration by capturing images at multiple known lens positions and pre-computing the relationship between phase difference measurements, lens defocus, and object distances. This calibration data is stored in advance and used during normal operation to accurately generate depth maps without requiring complex real-time calculations.

Inventive Principle:

Principle #10Preliminary action

Solution Approach 2:

The patent introduces calibration data as an intermediary that bridges the relationship between phase difference measurements and object distances. Instead of directly measuring object distance, the system uses phase difference measurements to determine lens defocus, then uses calibration data to translate lens defocus into object distance, enabling accurate depth map generation.

Inventive Principle:

Principle #24Intermediary (Mediator)

2Reliability

If calibration data is used to correlate phase difference with lens defocus, then depth map accuracy improves, but the system requires pre-stored calibration data and processing steps

Engineering Contradiction:

Improveautofocus performanceVSAvoidprocessing efficiency

Core Design Contradiction:

ReliabilityVSProductivity

Solution Approach 1:

The calibration process is performed in advance during system setup, where images are captured at multiple known lens positions and the relationship data is pre-computed and stored. During normal autofocus operation, the system simply retrieves pre-computed calibration data and performs straightforward lookups, significantly improving processing efficiency compared to real-time complex calculations.

Inventive Principle:

Principle #10Preliminary action

Solution Approach 2:

The system uses feedback from phase difference measurements to determine lens defocus, then uses this information along with calibration data to accurately determine object distance. This feedback loop ensures reliable autofocus performance by continuously adjusting based on measured phase differences while leveraging pre-computed calibration relationships.

Inventive Principle:

Principle #23Feedback

Data Source

PatentUS12546585B2Depth map construction by camera phase detection and lens calibration

Publication Date: 2026.02.10 ADVANCED MICRO DEVICES INC

AI summary

Techniques for generating a depth map are described. The techniques include obtaining a set of phase difference measurements with a phase detect sensor, and generating a depth map based on the set of phase difference measurements, utilizing a first set of calibration data correlating phase difference measurements with lens defocus data and a second set of calibration data correlating lens positions with object distances.