Target Point Specification Using Shape Model Fitting

Resolve Bottlenecks,
Find Innovative Solutions
Generate Solutions

Solution Overview

Problem

Existing technologies face challenges in accurately specifying target points of objects for calculating their position and speed, particularly in dynamic environments, due to limitations in object detection and tracking systems.

Innovation Solution

An information processing device equipped with a processing circuit, sensor, and output circuit that uses shape model management information to acquire, categorize, and fit shape models to detection points, thereby specifying a target point based on candidate points within the shape model.

Engineering Contradictions & Design Principles

VSEngineering Contradiction Analysis

1Device complexity

If a target point is calculated by simply selecting a detection point on the object, then the calculation process is simple, but the accuracy of specifying the target point deteriorates

Engineering Contradiction:
Improvecalculation process complexityVSAvoidtarget point specification accuracy
Core Design Contradiction:
Device complexityVSMeasurement precision

Solution Approach 1:

The patent pre-calculates and stores multiple candidate points on the object in advance, along with their relationship to the object's position and velocity. When tracking is needed, the system simply selects from these pre-prepared candidate points based on current detection data, avoiding complex real-time calculations while maintaining high accuracy in target point specification.

Inventive Principle:
Principle #10Preliminary action

2Device complexity

If conventional object detection and tracking systems are used, then the system structure is simple, but the reliability of target point specification deteriorates in dynamic environments

Engineering Contradiction:
Improvesystem structureVSAvoidtarget point specification reliability
Core Design Contradiction:
Device complexityVSReliability

Solution Approach 1:

The patent implements a feedback mechanism where the system continuously monitors the object's position and velocity, compares detected values with expected values based on the stored candidate points, and adjusts the selected target point accordingly. This feedback loop ensures reliable target point specification even in dynamic environments where object motion may cause detection variations.

Inventive Principle:
Principle #23Feedback

Data Source

PatentEP3252657B1Information processing device and information processing method
Publication Date: 2023.10.11 KK TOSHIBA
  • EP3252657B1 patent drawingFigure 1
  • EP3252657B1 patent drawingFigure 2
  • EP3252657B1 patent drawingFigure 3

AI summary

According to an arrangement, an information processing device includes an acquirer(20A), a categorizer(20F), a shape fitter(20G), and a specifier(20B). The acquirer is configured to acquire position information of each of a plurality of detection points. The categorizer(20F) is configured to cluster the detection points into one or more detection point groups each representing an object. The shape fitter(20G) is configured to apply a predetermined shape model to each detection point group on the basis of the position information of each of the clustered detection points included in the each detection point group. The specifier(20B) is configured to specify a target point of the object on the basis of the shape model.