3D data processing with advanced computer graphics tools

Abstract

Often, the 3-D raw data coming from an optical profilometer contains spiky noises and irregular grid, which make it difficult to analyze and difficult to store because of the enormously large size. This paper is to address these two issues for an optical profilometer by substantially reducing the spiky noise of the 3-D raw data from an optical profilometer, and by rapidly re-sampling the raw data into regular grids at any pixel size and any orientation with advanced computer graphics tools. Experimental results will be presented to demonstrate the effectiveness of the proposed approach. ABSTRACT Often, the 3-D raw data coming from an optical profilometer contains spiky noises and irregular grid, which make it difficult to analyze and difficult to store because of the enormously large size. This paper is to address these two issues for an optical profilometer by substantially reducing the spiky noise of the 3-D raw data from an optical profilometer, and by rapidly re-sampling the raw data into regular grids at any pixel size and any orientation with advanced computer graphics tools. Experimental results will be presented to demonstrate the effectiveness of the proposed approach.

Virtual tool mark generation for efficient striation analysis in forensic science

“…Moreover, the error is reduced by an entire order of magnitude when the number of samples used to represent the ideal geometry in the tool file is increased by a factor of ten. These results are similar to those found by Zhang et al (2012) for 2D re-sampling with OpenGL. Zhang et al (2012) found RMS values of a similar magnitude and observed the same trend of decreasing error with increasing resolution.…”

“…These results are similar to those found by Zhang et al (2012) for 2D re-sampling with OpenGL. Zhang et al (2012) found RMS values of a similar magnitude and observed the same trend of decreasing error with increasing resolution.…”

Optimization of a Statistical Algorithm for Objective Comparison of Toolmarks

“…To remove this noise, the authors used a combination of automated cleaning algorithms and manual cleaning. The automated cleaning algorithms are described in more detail in , but a brief description of these follows. First, the 2D image texture and quality map from the IFM operating software were used to remove those points that were too dark or had a poor quality value.…”

Development of a Mobile Toolmark Characterization/Comparison System

“…The MANTIS software is written in C++ and makes use of open‐source libraries and the following software: Open GL and a graphical processing unit to produce virtual marks of tools at any given angle and resolution and to visualize geometric data on the screen; QT to create the graphical user interface (GUI); and Java to interact with computers through command lines. While development of the basic software, consisting of the cleaning and data analysis routines, was carried out at AL/ISU , transforming the initial code into a more user‐friendly interface was achieved by working with Chris Hanson and Brian Bailey (Alphapixel, Evergreen, CO). A screenshot of the startup screen of the analysis suite is shown in Fig.…”