HDR Imaging via TDI Sensor Row Segmentation

Resolve Bottlenecks,
Find Innovative Solutions
Generate Solutions

Solution Overview

Problem

Conventional High Dynamic Range (HDR) imaging techniques struggle to effectively capture moving scenes due to the tradeoff between detector sensitivity and saturation, and are difficult to implement when the scene shifts between sequential acquisitions.

Innovation Solution

A method and system utilizing a time-delay integration (TDI) sensor with multiple pixel rows, each having different gain values and exposure times, which are dynamically adjusted and combined to produce an image with a high dynamic range in a single scan, allowing for improved HDR imaging even when the scene is moving.

Engineering Contradictions & Design Principles

VSEngineering Contradiction Analysis

1Illumination intensity

If conventional HDR techniques use sequential acquisitions at different exposure times, then dynamic range is improved, but implementation difficulty increases when the scene is moving

Engineering Contradiction:
Improvedynamic rangeVSAvoidimplementation difficulty
Core Design Contradiction:
Illumination intensityVSDevice complexity

Solution Approach 1:

The detector array is divided into multiple subarrays with different sensitivities, allowing each subarray to capture different brightness level ranges simultaneously. This segmentation enables HDR imaging without requiring sequential acquisitions, thus resolving the contradiction between improved dynamic range and implementation difficulty for moving scenes.

Inventive Principle:
Principle #1Segmentation

Solution Approach 2:

The patent transitions from temporal dimension (sequential acquisitions at different exposure times) to spatial dimension (multiple subarrays with different sensitivities captured simultaneously). This dimensional change allows HDR imaging of moving scenes by capturing multiple exposure levels in a single frame, eliminating the implementation difficulties associated with sequential methods.

Inventive Principle:
Principle #17Another dimension (Dimensionality change)

2Measurement precision

If detector sensitivity is increased to capture low signal, then low signal detection is improved, but detector saturation occurs with high signal

Engineering Contradiction:
Improvedetector sensitivityVSAvoiddetector saturation
Core Design Contradiction:
Measurement precisionVSObject-affected harmful factors

Solution Approach 1:

Different subarrays of the detector are assigned different sensitivities tailored to specific brightness level ranges. Low-sensitivity subarrays capture high signal without saturation, while high-sensitivity subarrays capture low signal with adequate precision. This local differentiation of detector properties resolves the contradiction between detector sensitivity and saturation.

Inventive Principle:
Principle #3Local quality

Solution Approach 2:

The patent changes the sensitivity parameter across different subarrays of the detector array, creating a multi-parameter detector system. By varying the sensitivity parameter spatially across the detector, the system can simultaneously optimize for both low signal detection and high signal capture without saturation.

Inventive Principle:
Principle #35Parameter changes

3Measurement precision

If automatic gain control algorithms are used to select tradeoff, then signal-to-noise ratio is improved, but system complexity increases

Engineering Contradiction:
Improvesignal-to-noise ratioVSAvoidsystem complexity
Core Design Contradiction:
Measurement precisionVSDevice complexity

Solution Approach 1:

The detector array performs self-service by having multiple subarrays with different sensitivities that automatically capture different brightness levels simultaneously. This eliminates the need for complex automatic gain control algorithms to select tradeoffs, as the system inherently captures multiple exposure levels in parallel, reducing system complexity while maintaining improved signal-to-noise ratio.

Inventive Principle:
Principle #25Self-service

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

Enables HDR imaging with improved dynamic range and reduced system costs, requiring less light, and is applicable to various applications including airborne and machine vision, by dynamically selecting gain values and exposure times for each pixel row, optimizing image acquisition across a wider range of irradiance.

Implementation Method 1

scanning with a time-delay integration (TDI) sensor to obtain image data from a first plurality of pixel rows and a second plurality of pixel rows

Methodology Applied
Scientific EffectTime-delay integration:

Data Source

PatentEP3336798B1HDR imaging
Publication Date: 2021.05.19 GOODRICH CORP
  • EP3336798B1 patent drawingFigure 1
  • EP3336798B1 patent drawingFigure 2

AI summary

The present disclosure generally relates to HDR imaging techniques, and more specifically to HDR imaging techniques for use when a scene is moving. For time delay integration, the same scene location is repeatedly imaged on sequential rows, allowing for different gain values and/or exposure times to be utilized in different rows. The present disclosure utilizes a static or dynamic selection of gain values and/or exposure times on each row to enable stitching of the rows for high dynamic range.