Browse Prior Art Database

METHOD FOR ANALYZING COLORS ON A THREE DIMENSIONAL (3D) SURFACE

IP.com Disclosure Number: IPCOM000244889D
Publication Date: 2016-Jan-27
Document File: 4 page(s) / 70K

Publishing Venue

The IP.com Prior Art Database

Abstract

A technique of automated color segmentation and analysis for images of machine parts on which thermal paint is applied is disclosed. The machine part that requires analysis is placed on a rotary motion stage or a robot manipulator. The stage or robot manipulates the machine part into a number of positions for image collection. The positions can be obtained through a variety of techniques, including pre-determination using a three dimensional (3D) modeling software. Further, viewing positions are established such that when images are collected from the different viewing positions, each area of interest on the part is viewed in at least one of the images from a normal, or near-normal viewing angle. After image collection, the images are mapped to a 3D surface. The color assigned to each 3D point of interest is then taken from an image that viewed the point from the angle nearest to normal. If multiple images fit such a category, then pixel value of such images may be averaged or otherwise aggregated for the 3D point.

This text was extracted from a Microsoft Word document.
At least one non-text object (such as an image or picture) has been suppressed.
This is the abbreviated version, containing approximately 45% of the total text.

METHOD FOR ANALYZING COLORS ON A THREE DIMENSIONAL (3D) SURFACE

BACKGROUND

The present invention relates generally to machine parts of a combustion system that undergo temperature excursions and more particularly to a technique for automated color segmentation and analysis of images of machine parts on which thermal paint is applied.

Generally, thermal paint is used for assessing thermal information from machine parts that undergo temperature excursions. Thermal paint is applied on machine part and change of color of the paint with temperature is analyzed. The paint changes color based on the highest temperature to which the paint is exposed. The machine part is subsequently manually analyzed to determine distribution of temperatures across the machine part. However, manual examination is time consuming and error-prone.

A conventional technique includes automated approach to assess calibrated temperatures from images of parts that are coated with the thermal paint. Automated assessment of temperature change of the machine part is faster and more reliable. However, such an approach suffers from errors in cases where geometry of the machine part is such that the machine part is required to be viewed from a variety of angles. If the part has areas having high curvature, an imaging device is not able to view every portion of the machine part at normal incidence. When an object is viewed from a variety of angles, same patch on the machine part may have appeared different depending on bidirectional reflectance distribution function (BDRF) of surface that is viewed. When colors are imaged from disparate viewpoints, the same patch may have different color appearance depending on the viewpoint. Such an approach is problematic if one-to-one correspondence between color and temperature is established.

It would be desirable to have a technique to provide an automated approach for image analysis of a machine part from a variety of viewing angles and take into account BDRF of the surface being viewed.

BRIEF DESCRIPTION OF DRAWINGS

Figure 1 depicts 3D model of a machine part that is analyzed.

Figure 2 depicts image captured for 3D model of the machine part that is imaged for analysis.

 

Figure 3 depicts computed viewpoint angles (degrees) for 3D model of the machine part that is imaged for analysis.

DETAILED DESCRIPTION

A technique of automated color segmentation and analysis for images of machine parts on which thermal paint is applied is disclosed. The machine part that requires analysis is placed on a rotary motion stage or a robot manipulator. The stage or robot manipulates the machine part into a number of positions for image collection. The positions can be obtained through a variety of techniques, including pre-determination using a three dimensional (3D) modeling software. Alternatively, viewpoint of the machine part can be established in multiple views by identifying manually or automatically features on a 3D model of the machine part and identify...