Passive Range Finder Using AI Image Segmentation

Resolve Bottlenecks,
Find Innovative SolutionsGenerate Solutions

Solution Overview

Problem

Existing ranging methods are either traceable or inaccurate, with laser-based systems susceptible to tracing and human error, and stadia metrics lacking flexibility and accuracy.

Innovation Solution

A passive ranging system using artificial intelligence models for object segmentation and pose estimation to determine the range to a target without active sensing, capable of handling moving and non-upright targets, and incorporating focal length information for precise calculations.

Engineering Contradictions & Design Principles

VSEngineering Contradiction Analysis

1Reliability

If laser-based active ranging systems are used, then ranging capability is provided, but the system becomes traceable and susceptible to detection

Engineering Contradiction:

Improveranging capabilityVSAvoidtraceability

Core Design Contradiction:

ReliabilityVSObject-affected harmful factors

Solution Approach 1:

The patent inverts the conventional active ranging approach by using a passive system that analyzes target dimensions in images rather than transmitting signals to the target. Instead of sending out laser beams or signals and measuring reflections, the system captures images and uses AI to estimate target size and calculate range based on apparent dimensions, making the ranging process undetectable and untraceable.

Inventive Principle:

Principle #13The other way round (Inversion)

2Ease of operation

If stadia metrics are used for ranging, then simplicity is achieved, but accuracy and flexibility are reduced

Engineering Contradiction:

Improveranging simplicityVSAvoidranging accuracy

Core Design Contradiction:

Ease of operationVSMeasurement precision

Solution Approach 1:

The patent replaces traditional mechanical stadia metrics with an AI-based image analysis system. Instead of using physical reticles and manual measurements, the system uses neural networks to automatically detect target boundaries, estimate dimensions, and calculate range, significantly improving accuracy while maintaining ease of operation through automated processing.

Inventive Principle:

Principle #28Mechanics substitution (Replace mechanical system)

Solution Approach 2:

The system changes the parameters used for ranging by transitioning from fixed stadia measurements to dynamic AI-based dimension estimation. The AI model analyzes multiple image parameters including target boundary coordinates, pixel dimensions, and contextual information to calculate accurate range, providing flexibility for different target types and conditions.

Inventive Principle:

Principle #35Parameter changes

3Adaptability or versatility

If human operators perform ranging, then flexibility is provided, but human error reduces accuracy

Engineering Contradiction:

Improveranging flexibilityVSAvoidranging accuracy

Core Design Contradiction:

Adaptability or versatilityVSMeasurement precision

Solution Approach 1:

The patent implements a self-service ranging system where the AI model automatically performs target detection, dimension estimation, and range calculation without human intervention. The system serves itself by processing images through neural networks that identify target boundaries and compute range based on apparent dimensions, eliminating human error while maintaining flexibility through automated adaptation to different target types.

Inventive Principle:

Principle #25Self-service

4Object-affected harmful factors

If passive image-based ranging is used, then untraceability is achieved, but handling partial obstruction and non-upright poses becomes difficult

Engineering Contradiction:

ImproveuntraceabilityVSAvoidtarget condition handling

Core Design Contradiction:

Object-affected harmful factorsVSAdaptability or versatility

Solution Approach 1:

The patent applies preliminary action by training the AI model on diverse datasets that include partially obscured targets and non-upright poses before deployment. The neural network learns to handle various target conditions in advance, enabling it to accurately estimate dimensions and calculate range even when targets are partially blocked or in unusual positions, thus improving adaptability while maintaining passive untraceable operation.

Inventive Principle:

Principle #10Preliminary action

Data Source

PatentUS12450762B2Passive range finder and associated method

Publication Date: 2025.10.21 QIOPTIQ LTD

AI summary

Described herein are passive systems and methods for ranging objects. The systems and methods are passive in that they do not rely on transmission of radiation (whether radiofrequency, visible, infrared, ultraviolet, acoustic, etc.) to determine range. The systems and methods described herein may be used both in civilian and military applications. The present techniques allow users to determine range to moving targets, which is notoriously difficult to estimate using the human eye. The present techniques perform ranging to a target using artificial intelligence models, including object segmentation and pose estimation. Object segmentation may involve determining a dimension (e.g., height) of a specified target in an image, in terms of pixels. If there are multiple objects in the scene, range estimates may be calculated using object segmentation for each object. Pose estimation involves a machine learning technique that identifies sets of coordinates corresponding to joint keypoints.