Trajectory-Based Sensor Capture for Selective Image Processing

Resolve Bottlenecks,
Find Innovative Solutions
Generate Solutions

Solution Overview

Problem

Existing mobile vehicle systems, such as trains equipped with image, LIDAR, and RADAR sensors, face challenges in data usability due to contradictory data, exposure issues in tunnels, and resource-intensive processing, particularly in adverse weather conditions like fog or dust.

Innovation Solution

A method that limits data analysis to specific subsections of interest, optimizing sensor parameters and processing resources by using a control block to determine and adjust image capture and processing parameters based on predefined trajectory points and associated information, focusing effort on areas relevant for obstacle detection and safety.

Engineering Contradictions & Design Principles

VSEngineering Contradiction Analysis

1Measurement precision

If data from the entire acquisition area is processed, then complete environmental coverage is achieved, but processing time and computing resources increase significantly

Engineering Contradiction:
Improveenvironmental coverageVSAvoidprocessing time
Core Design Contradiction:
Measurement precisionVSLoss of time

Solution Approach 1:

The image is divided into multiple sub-sections corresponding to different trajectory points. The control block identifies which sub-sections are relevant to the predefined trajectory and processes only those, rather than analyzing the entire image. This segmentation approach maintains necessary environmental coverage while significantly reducing processing time and computing resource requirements.

Inventive Principle:
Principle #1Segmentation

2Illumination intensity

If sensor exposure is increased to capture tunnel environments, then visibility in dark areas improves, but overexposure occurs in bright areas creating black halos

Engineering Contradiction:
Improvetunnel visibilityVSAvoiddata usability
Core Design Contradiction:
Illumination intensityVSReliability

Solution Approach 1:

Different exposure parameters are applied to different sub-sections of the image based on their location relative to the trajectory. The control block determines optimal capture parameters for each sub-section, allowing localized exposure adjustment. This enables proper illumination in tunnel areas without causing overexposure in bright areas, maintaining data usability across the entire scene.

Inventive Principle:
Principle #3Local quality

3Measurement precision

If all capture parameters are optimized for the entire acquisition area, then overall image quality improves, but processing complexity and resource requirements increase

Engineering Contradiction:
Improveimage qualityVSAvoidprocessing complexity
Core Design Contradiction:
Measurement precisionVSDevice complexity

Solution Approach 1:

The control block dynamically determines capture parameters for each sub-section based on its specific characteristics and relevance to the predefined trajectory. Rather than applying a static uniform optimization to the entire image, the system adapts parameters locally for each region, achieving high image quality where needed while reducing processing complexity through selective optimization.

Inventive Principle:
Principle #15Dynamics

4Quantity of substance

If comprehensive data acquisition is performed across all sensor areas, then complete environmental monitoring is achieved, but data usability decreases due to contradictory and irrelevant information

Engineering Contradiction:
Improvedata volumeVSAvoiddata usability
Core Design Contradiction:
Quantity of substanceVSReliability

Solution Approach 1:

The control block extracts and processes only the sub-sections of the image that are relevant to the predefined trajectory and safety-critical areas. Irrelevant portions of the acquisition area are excluded from processing. This extraction approach maintains comprehensive monitoring of critical regions while eliminating contradictory and irrelevant information that reduces data usability.

Inventive Principle:
Principle #2Taking out (Extraction)

Data Source

PatentEP3716097B1Method for acquiring data captured by a capture module embedded in a mobile device following a predetermined trajectory, corresponding computer program and device
Publication Date: 2024.08.14 ALSTOM HOLDINGS SA
  • EP3716097B1 patent drawingFigure 1
  • EP3716097B1 patent drawingFigure 2
  • EP3716097B1 patent drawingFigure 3

AI summary

A method for acquiring data, in an electronic acquisition module, captured by a sensor mounted on a vehicle with a predefined trajectory, comprising the following steps implemented in real time by an electronic acquisition module comprising a database defining, for each of a plurality of points on the predefined trajectory, a set of subsections of interest indicating at least one subsection of interest (30) in the acquisition area of ​​the sensor: - determination of a next position of the mobile machine on said predefined trajectory of the mobile machine; - determination, as a function of at least the subsection of interest (30) defined by the database for the point corresponding to the determined position, of the value of a parameter including data acquisition by the sensor mounted at said next position or of a parameter for processing the data captured by the sensor mounted at said next position.