Stereo Camera Distance Measurement via Variable Image Resolution

Resolve Bottlenecks,
Find Innovative Solutions
Generate Solutions

Solution Overview

Problem

Existing stereo camera systems face challenges in achieving high accuracy and stability for distance measurement, particularly at distant locations, as they apply uniform processing methods that increase computational complexity without effectively improving distance resolution for both near and far objects.

Innovation Solution

The system dynamically adjusts the resolution of specific image areas based on the distance of objects, using image generating units to correct distortions and increase resolution only in distant areas, allowing for accurate disparity calculation without altering matching block sizes.

Engineering Contradictions & Design Principles

VSEngineering Contradiction Analysis

1Measurement precision

If uniform high-resolution processing is applied to the entire image to improve distance resolution for distant objects, then distance measurement accuracy for distant objects improves, but computational complexity increases excessively

Engineering Contradiction:
Improvedistance measurement accuracyVSAvoidcomputational complexity
Core Design Contradiction:
Measurement precisionVSDevice complexity

Solution Approach 1:

The patent applies different processing resolutions to different regions of the image based on object distance. Specifically, the image is divided into a first region (distant objects) and a second region (near objects), with the first region processed at high resolution and the second region processed at low resolution. This local differentiation resolves the contradiction by providing high measurement precision only where needed (distant objects) while avoiding excessive computational complexity in regions where it is not needed (near objects).

Inventive Principle:
Principle #3Local quality

Solution Approach 2:

The patent segments the image processing task into distinct regions based on object distance. The processing unit divides the image into multiple regions and applies different disparity calculation methods to each region. This segmentation allows the system to optimize computational resources by applying complex high-resolution processing only to distant objects while using simpler low-resolution processing for near objects, thereby resolving the contradiction between measurement precision and computational complexity.

Inventive Principle:
Principle #1Segmentation

2Measurement precision

If matching block size is reduced to improve distance resolution for distant objects, then distance measurement stability improves, but the number of calculations increases

Engineering Contradiction:
Improvedistance measurement stabilityVSAvoidprocessing speed
Core Design Contradiction:
Measurement precisionVSProductivity

Solution Approach 1:

The patent applies different matching block sizes to different regions of the image based on object distance. For distant objects in the first region, smaller matching blocks are used to improve distance measurement stability. For near objects in the second region, larger matching blocks are used to reduce the number of calculations. This local differentiation resolves the contradiction by optimizing measurement stability only where needed while maintaining processing speed in regions where it is not compromised.

Inventive Principle:
Principle #3Local quality

Solution Approach 2:

The patent segments the image into regions with different object distance characteristics and applies different matching block sizes to each segment. This segmentation strategy allows the system to use computationally intensive small matching blocks only for distant objects where stability is critical, while using larger matching blocks for near objects where processing speed is more important, thereby resolving the contradiction between measurement stability and processing speed.

Inventive Principle:
Principle #1Segmentation

3Measurement precision

If image resolution is increased for distant areas only, then distance resolution for distant objects improves, but device complexity increases

Engineering Contradiction:
Improvedistance resolutionVSAvoidprocessing complexity
Core Design Contradiction:
Measurement precisionVSDevice complexity

Solution Approach 1:

The patent implements local quality enhancement by increasing image resolution only in the first region (distant objects) while maintaining standard resolution in the second region (near objects). The processing unit identifies distant objects and applies high-resolution processing selectively to their corresponding image regions. This resolves the contradiction by providing improved distance resolution only where needed for distant objects while avoiding the device complexity of processing the entire image at high resolution.

Inventive Principle:
Principle #3Local quality

Data Source

PatentEP2546602B1Stereo camera apparatus
Publication Date: 2017.09.13 HITACHI AUTOMOTIVE SYST LTD
  • EP2546602B1 patent drawingFigure 1
  • EP2546602B1 patent drawingFigure 2(a)~2(b)
  • EP2546602B1 patent drawingFigure 3(a)~3(b)

AI summary

A stereo camera apparatus which carries out distance measuring stably and with high accuracy by making measuring distance resolution variable according to a distance to an object is provided. A stereo camera apparatus 1 takes in two images, changes resolution of a partial area of each image that is taken in, and calculates a distance from a vehicle to an object that is imaged in the partial area, based on disparity of the partial area of each image in which resolution is changed. Thus, even when the object exists at a long distance and is small in size, distance measuring processing can be carried out stably.