Pose Estimation via Critical Point Analysis
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Conventional pose estimation techniques for machine vision are complex and expensive, requiring specialized 3D camera systems and marker systems that restrict subject movement and preparation, making them impractical for widespread use.
Innovation Solution
A system that uses depth information from a camera to perform critical point analysis, segmenting images, identifying local minimums and maximums, and connecting these points to estimate subject poses, allowing for the detection of joint sections and calculating probabilities to determine object orientation without the need for expensive equipment.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Measurement precision
If conventional 3D camera systems are used for pose estimation, then measurement precision is improved, but device complexity and cost increase significantly
Solution Approach 1:
The patent uses 2D images as copies or projections of the 3D scene, extracting pose information from 2D image data rather than requiring direct 3D measurement. The algorithm reconstructs 3D pose from multiple 2D views, effectively using 2D copies to represent 3D information.
Solution Approach 2:
The patent replaces complex mechanical 3D camera systems with a computational approach using standard 2D cameras and image processing algorithms. The mechanical complexity of 3D sensing is substituted with algorithmic processing of 2D images.
2Measurement precision
If marker systems are used for pose estimation, then measurement precision is improved, but ease of operation deteriorates due to preparation requirements
Solution Approach 1:
The patent enables the subject to serve itself for pose estimation without external markers. The algorithm automatically detects and tracks body parts using natural image features, allowing the subject to be observed without any preparation or cooperation beyond being visible in the image.
3Measurement precision
If conventional pose estimation techniques are used, then measurement precision is improved, but device complexity increases
Solution Approach 1:
The patent extracts pose estimation capability from complex dedicated equipment and implements it as a software algorithm running on standard computing hardware. The functionality is extracted from specialized devices and embedded in a general-purpose computational system.
Data Source
AI summary
Methods and systems for estimating a pose of a subject. The subject can be a human, an animal, a robot, or the like. A camera receives depth information associated with a subject, a pose estimation module to determine a pose or action of the subject from images, and an interaction module to output a response to the perceived pose or action. The pose estimation module separates portions of the image containing the subject into classified and unclassified portions. The portions can be segmented using k-means clustering. The classified portions can be known objects, such as a head and a torso, that are tracked across the images. The unclassified portions are swept across an x and y axis to identify local minimums and local maximums. The critical points are derived from the local minimums and local maximums. Potential joint sections are identified by connecting various critical points, and the joint sections having sufficient probability of corresponding to an object on the subject are selected.