Adaptive High Dynamic Range Camera Using Super Pixel Segmentation

Resolve Bottlenecks,
Find Innovative Solutions
Generate Solutions

Solution Overview

Problem

Time of flight (TOF) 3D cameras face challenges in achieving a dynamic range sufficient for accurate distance measurements due to shot and read noise, especially for features with low reflectivity or at large distances, resulting in unacceptably large errors and an inability to capture both dim and bright features effectively.

Innovation Solution

The camera employs a photosensor with light-sensitive pixels tiled into super pixels, allowing for improved dynamic range by subdividing the photosurface into areas that group pixels for enhanced reading, thereby increasing the minimum and maximum integrated irradiance, leading to a camera dynamic range that is about N times that of a conventional camera.

Engineering Contradictions & Design Principles

VSEngineering Contradiction Analysis

1Measurement precision

If camera parameters and lighting are adjusted to compensate for low reflectivity and large distances of features in the scene, then measurement precision for dim features is improved, but near and high reflectivity features cause pixel saturation leading to loss of information

Engineering Contradiction:
Improvedistance measurement precision for dim featuresVSAvoidlight information from bright features
Core Design Contradiction:
Measurement precisionVSLoss of information

Solution Approach 1:

The photosensor is divided into multiple pixels, and the patent implements adaptive exposure control where each pixel can have different exposure parameters. This segmentation allows dim features to be captured with longer exposure while bright features use shorter exposure, preventing saturation and information loss.

Inventive Principle:
Principle #1Segmentation

Solution Approach 2:

The camera system dynamically adjusts exposure parameters based on the brightness characteristics of different regions in the scene. The patent describes adaptive exposure control that modifies exposure time and gain settings in real-time to accommodate varying light intensities across different features, ensuring optimal measurement precision without saturation.

Inventive Principle:
Principle #15Dynamics

2Measurement precision

If the exposure period is extended to increase the amount of gated light registered by pixels, then measurement precision for distant features is improved, but pixels saturate more easily causing loss of information for brighter features

Engineering Contradiction:
Improvelight measurement precision for distant featuresVSAvoidlight information from bright features
Core Design Contradiction:
Measurement precisionVSLoss of information

Solution Approach 1:

The patent implements dynamic exposure control that adjusts the exposure period based on the brightness of the imaged features. For distant, dim features, the exposure period is extended to accumulate sufficient photoelectrons for accurate measurement. For brighter features, the exposure period is reduced to prevent pixel saturation and information loss.

Inventive Principle:
Principle #15Dynamics

Solution Approach 2:

Different regions of the photosensor can have different exposure parameters applied locally. The patent describes adaptive exposure control that tailors exposure settings to the specific brightness characteristics of each region or feature being imaged, allowing optimal exposure for both dim and bright areas simultaneously.

Inventive Principle:
Principle #3Local quality

3Measurement precision

If the number of photoelectrons accumulated in pixels is increased to reduce shot noise error, then measurement precision is improved, but the dynamic range of the camera is reduced causing pixel saturation

Engineering Contradiction:
Improvelight measurement precisionVSAvoidcamera dynamic range
Core Design Contradiction:
Measurement precisionVSAdaptability or versatility

Solution Approach 1:

The photosensor is segmented into multiple pixels with independent exposure control. This allows the system to accumulate sufficient photoelectrons in pixels imaging dim features to reduce shot noise, while pixels imaging bright features use different accumulation parameters to avoid saturation, thereby maintaining overall dynamic range.

Inventive Principle:
Principle #1Segmentation

Solution Approach 2:

The camera dynamically adjusts the number of photoelectrons accumulated based on the brightness of the imaged features. For dim features, the system allows longer accumulation to reduce shot noise and improve precision. For bright features, accumulation is limited to prevent saturation, preserving the camera's dynamic range across different scene conditions.

Inventive Principle:
Principle #15Dynamics

Applied Scientific Principles

This section explains which scientific principles are used to turn an abstract innovation direction into a practical engineering solution.

Function Achieved in This Case

This approach significantly reduces measurement errors by increasing the dynamic range, allowing for more accurate distance measurements with improved precision and the ability to capture a wider range of light intensities, resulting in a camera that can effectively image both dim and bright features.

Implementation Method 1

A pixel in the photosensor registers an amount of incident light it receives by generating, and accumulating, a quantity of electrons, hereinafter 'photoelectrons', in the pixel that is substantially proportional to the amount of incident light

Methodology Applied
Scientific EffectPhotoelectric effect: Photoelectric Effect

Data Source

PatentUS9137463B2Adaptive high dynamic range camera
Publication Date: 2015.09.15 MICROSOFT TECHNOLOGY LICENSING LLC
  • US9137463B2 patent drawing
  • US9137463B2 patent drawing
  • US9137463B2 patent drawing

AI summary

An embodiment of the invention provides a time of flight 3D camera comprising a photosensor having a plurality of pixels that generate and accumulate photoelectrons responsive to incident light, which photosensor is tiled into a plurality of super pixels, each partitioned into a plurality of pixel groups and a controller that provides a measure of an amount of light incident on a super pixel responsive to quantities of photoelectrons from pixel groups in the super pixel that do not saturate a readout pixel comprised in the photosensor.