Image Content Engine (ICE): A System for Fast Image Database Searches Page: 3 of 7
This article is part of the collection entitled: Office of Scientific & Technical Information Technical Reports and was provided to Digital Library by the UNT Libraries Government Documents Department.
The following text was automatically extracted from the image on this page using optical character recognition software:
Image Content Engine (ICE): A System
for Fast Image Database Searches
James M. Brase, Douglas N. Poland, David W. Paglieroni, George F. Weinert,
Charles W. Grant, Aseneth S. Lopez and Sergei Nikolaev
Lawrence Livermore National Laboratory, P.O. Box 808, L-210, Livermore, CA 94551
The Image Content Engine (ICE) is being developed to provide cueing assistance to human image analysts faced with
increasingly large and intractable amounts of image data. The ICE architecture includes user configurable feature
extraction pipelines which produce intermediate feature vector and match surface files which can then be accessed by
interactive relational queries. Application of the feature extraction algorithms to large collections of images may be
extremely time consuming and is launched as a batch job on a Linux cluster. The query interface accesses only the
intermediate files and returns candidate hits nearly instantaneously. Queries may be posed for individual objects or
collections. The query interface prompts the user for feedback, and applies relevance feedback algorithms to revise the
feature vector weighting and focus on relevant search results. Examples of feature extraction and both model-based and
search-by-example queries are presented.
Keywords: broad area search, model matching, image search, image analysis
New sensor and communication technologies provide scientists and analysts with a deluge of data in many forms:
images, simulations, experimental measurements, communications intercepts, and textual data. Extracting knowledge
from these masses of data requires the ability to discover subtle correlations in complex data sets - correlations that
indicate relationships that lead to understanding
The Image Content Engine (ICE) is a set of software tools which allow an analyst to search a large set of images for a
specific type of object. ICE provides two approaches for specifying the search: the target can be specified by a three-
dimensional model or it can be specified by a set of example images. ICE provides image pipeline tools for extracting
image features and matching object models, interactive relevance feedback tools for providing feedback on search
performance, and tools for relational queries of the resulting detection databases.
The specific targets of ICE are very large image sets - terabytes in many cases. The feature extraction and object
detection pipelines are implemented on parallel Linux clusters. A toolbox for mapping image processing pipelines to
clusters and managing associated metadata with XML has been developed. The ICE system has been demonstrated on
small clusters and is being scaled to larger systems for specific applications. In this paper we will describe the ICE
system architecture and give examples of model-based target search and search-by-example applications.
2. THE ICE ARCHITECTURE
The top-level architecture of ICE is shown in Figure 1. We begin at the upper left with a set of images that is to be
searched. The source of images can be either a real-time sensor stream or a pre-existing database. Image features and
target matching calculations are performed in a set of image processing pipelines shown on the left side of the Figure.
The tile feature pipeline computes a pre-defined set of features for each image tile - small (typically 128 pixels on a
side) regularly-spaced overlapping rectangular areas. The object matching pipeline performs phase-sensitive image
matching with a predefined user-selected set of three-dimensional object models. Extracted image features and target
matching metrics are stored in a database.
Here’s what’s next.
This article can be searched. Note: Results may vary based on the legibility of text within the document.
Tools / Downloads
Get a copy of this page or view the extracted text.
Citing and Sharing
Basic information for referencing this web page. We also provide extended guidance on usage rights, references, copying or embedding.
Reference the current page of this Article.
Brase, J M; Paglieroni, D W; Weinert, G F; Grant, C W; Lopez, A S & Nikolaev, S. Image Content Engine (ICE): A System for Fast Image Database Searches, article, March 22, 2005; Livermore, California. (https://digital.library.unt.edu/ark:/67531/metadc887962/m1/3/: accessed March 26, 2019), University of North Texas Libraries, Digital Library, https://digital.library.unt.edu; crediting UNT Libraries Government Documents Department.