Visualization and Intelligent Systems Laboratory
VISLab

 

 

Contact Information

VISLab
Winston Chung Hall Room 216
University of California, Riverside
900 University Avenue
Riverside, CA 92521-0425


Tel: (951)-827-3954

CRIS
Bourns College of Engineering
UCR
NSF IGERT on Video Bioinformatics

UCR Collaborators:
CSE
ECE
ME
STAT
PSYC
ENTM
BIOL
BPSC
ECON
MATH
BIOENG
MGNT

Other Collaborators:
Keio University

Other Activities:
IEEE Biometrics Workshop 2014
IEEE Biometrics Workshop 2013
Worshop on DVSN 2009
Multibiometrics Book

Webmaster Contact Information:
Michael Caputo
michael.vislab@gmail.com

Last updated: June 15, 2016

 

 

Image Segmentation

Codebook Optimization using Word Activation Forces for Scene Categorization

Visual codebook based quantization of robust appearance descriptors extracted from local image patches is an effective means of capturing image statistics for texture analysis and natural scene classification. Based on the newly proposed statistics of word activation forces (WAFs), we have optimized the codebook. Currently, codebooks are typically created from a set of training images using a clustering algorithm. However, these codebooks are often functionally limited due to redundancy. We show that WAFs can remove the redundancy efficiently. In the experiment, the proposed method achieved the state-of-the-art performance on the Caltech-101, fifteen natural scene categories and VOC2007 databases. The optimization method also offers insights into the success of several recently proposed images classification approaches, including vector quantization (VQ) coding in the Spatial Pyramid Matching (SPM), sparse coding SPM (ScSPM), and Locality-constrained Linear Coding (LLC).

Reflection Symmetry-Integrated Image Segmentation

Examples of symmetry-integrated segmentation results usingimages from the Caltech-101 database We developed a new symmetry-integrated region-based image segmentation method to obtain improved image segmentation. The method constructs a symmetry token that can be flexibly embedded into segmentation cues. The method has been investigated experimentally in challenging natural images and images containing man-made objects. It is shown that the proposed method outperforms current segmentation methods both with and without exploiting symmetry. A thorough experimental analysis indicates that symmetry plays an important role as a segmentation cue, in conjunction with other attributes like color and texture.

Symmetry Integrated Region-based Image Segmentation

Symmetry is an important cue for machine perception that involves high-level knowledge of image components. Unlike most of the previous research that only computes symmetry in an image, this research integrates symmetry with image segmentation to improve the segmentation performance. The symmetry integration is used to optimize both the segmentation and the symmetry of regions simultaneously. Interesting points are initially extracted from an image and they are further refined for detecting symmetry axis. A symmetry affinity matrix is used explicitly as a constraint in a region growing algorithm in order to refine the symmetry of segmented regions. Experimental results and comparisons from a wide domain of images indicate a promising improvement by symmetry integrated image segmentation compared to other image segmentation methods that do not exploit symmetry.

Functional Template-based SAR Image Segmentation

We present an approach to automatic image segmentation, in which user selected sets of examples and counter-examples supply information about the specific segmentation problem. In our approach, image segmentation is guided by a genetic algorithm which learns the appropriate subset and spatial combination of a collection of discriminating functions, associated with image features. The genetic algorithm encodes discriminating functions into a functional template representation, which can be applied to the input image to produce a candidate segmentation.


Synthesizing Feature Agents Using Evolutionary Computation

We used genetic programming (GP) with smart crossover and smart mutation to discover integrated feature agents that are evolved from combinations of primitive image processing operations to extract regions-of-interest (ROIs) in remotely sensed images. Smart crossover and smart mutation identify and keep the effective components of integrated operators called "agents" and significantly improve the efficiency of GP. Our experimental results show that compared to normal GP, our GP algorithm with smart crossover and smart mutation can find good agents more quickly during training to effectively extract the regions-of-interest and the learned agents can be applied to extract ROIs in other similar images.

Adaptive Integrated Image Segmentation and Object Recognition

We present a general approach to image segmentation and object recognition that can adapt the image segmentation algorithm parameters to the changing environmental conditions. The edge-border coincidence measure is first used as reinforcement for segmentation evaluation to reduce computational expenses associated with model matching during the early stage of adaptation. This measure alone, however, can not reliably predict the outcome of object recognition. Therefore, it is used in conjunction with model matching. Results are presented for both indoor and outdoor color images where the performance improvement over time is shown for both image segmentation and object recognition.

Delayed Reinforcement Learning for Adaptive Image Segmentation

Generally, object recognition systems are open loop with no feedback between levels and assuring their robustness is a key challenge in computer vision and pattern recognition research. A robust closed-loop system based on “delayed” reinforcement learning is introduced in this paper. The parameters of a multilevel system employed for model-based object recognition are learned. The approach systematically controls feedback in a multilevel vision system and shows promise in approaching a long-standing problem in the field of computer vision and pattern recognition.

Learning Integrated Image Segmentation and Object Recognition

A general approach to image segmentation and object recognition that learned a mapping from images with varying properties to segmentation algorithm parameters is presented. The mapping was built using a reinforcement learning algorithm that was based on a team of generalized stochastic learning automata and operated separately in a global or local manner on an image. The edge-border coincidence was first used as an immediate reinforcement to reduce computational expenses associated with model matching during the early stage of the learning process. Since this measure could not reliably predict the outcome of object recognition, it was used in conjunction with model matching that provided optimal segmentation evaluation in a closed-loop object recognition system. Results are presented for both indoor and outdoor color images where the performance improvement over time is shown for both image segmentation and object recognition.

Analysis of Terrain using Multispectral images

Automated terrain analysis was required for many practical applications, such as outdoor navigation, image exploitation, remote sensing, reconnaissance and surveillance. A hierarchical approach to analyze multispectral (MS) imagery for autonomous land vehicle navigation is presented. The approach integrated several strategies to label various terrain classes in these images acquired using twelve spectral bands in the visible and near-infrared spectrum. At the low (pixel) level, it combined texture gradient results from specifically selected channels by varying the size of gradient operators and performed multi-thresholding and relaxation-based edge linking operations to obtain robust closed region boundaries. At the high (symbolic) level, it made use of the spectral, locational, and relational constraints among regions to achieve accurate terrain image interpretation.

Adaptive Image Segmentation Using Genetic and Hybrid Search Methods

An adaptive approach for the important image processing problem of image segmentation that relied on learning from experience to adapt and improve the Segmentation performance. The adaptive image segmentation system incorporated a feedback loop consisting of a machine learning subsystem, an image segmentation algorithm, and an evaluation component which determined segmentation quality. The machine learning component was based on genetic adaptation and used (separately) a pure genetic algorithm (GA) and a hybrid of GA and Hill Climbing (HC). When the learning subsystem was based on pure genetics, the corresponding evaluation component was based on a vector of evaluation criteria. For the hybrid case, the system employed a scalar evaluation measure which was a weighted combination of the different criteria. The multi-objective optimization demonstrated the ability of the adaptive image segmentation system to provide high quality segmentation results in a minimal number of generations.

Adaptive Image Segmentation Using a Genetic Algorithm

One of the fundamental weaknesses of computer vision systems used in practical applications was their inability to adapt the segmentation process as real-world changes occurred in the image. Presented is the first closed loop image segmentation system which incorporated a genetic algorithm to adapt the segmentation process to changes in image characteristics caused by variable environmental conditions such as time of day, time of year, clouds, etc. The segmentation problem was formulated as an optimization problem and the genetic algorithm efficiently searched the hyperspace of segmentation parameter combinations to determine the parameter set which maximized the segmentation quality criteria. The goals of the adaptive image segmentation system were to provide continuous adaptation to normal environmental variations, to exhibit learning capabilities, and to provide robust performance when interacting with a dynamic environment. Also presented are experimental results which demonstrated learning and the ability to adapt the segmentation performance in outdoor color imagery.

Adaptive Image Segmentation Using Multi-Objective Evaluation and Hybrid Search Methods

An approach for image segmentation that relied on learning from experience to adapt and improve the segmentation performance is presented. The adaptive image segmentation system incorporated a feedback loop that consisted of a machine learning subsystem, an image segmentation algorithm, and an evaluation component which determined segmentation quality. The machine learning component was based on genetic adaptation and used (separately) a pure genetic algorithm (GA) and a hybrid of GA and hill climbing (HC). When the learning subsystem was based on pure genetics, the corresponding evaluation component was based on a vector of evaluation criteria. For the hybrid case, the system employed a scalar evaluation measure which was a weighted combination of the different criteria. Experimental results for pure genetic and hybrid search methods are presented using a representative database of outdoor TV imagery.

Self-Optimizing Image Segmentation System Using a Genetic Algorithm

One of the fundamental weaknesses of computer vision systems used in outdoor applications was their inability to adapt the segmentation process as real-world changes occurred in the image. We present a closed loop image segmentation system that incorporated a genetic algorithm to adapt the segmentation process to changes in image characteristics caused by variable environmental conditions. The genetic algorithm efficiently searched the hyperspace of segmentation parameter combinations to determine the parameter set which maximized the segmentation quality criteria. A summary of the experimental results that demonstrated the ability to perform adaptive image segmentation and to learn from experience using a collection of outdoor color imagery is presented.

Closed-Loop Adaptive Image Segmentation

One of the fundamental weaknesses of computer vision systems used in outdoor applications was their inability to adapt the segmentation process as real-world changes occurred in the image. We present a closed loop image segmentation system that incorporated a genetic algorithm to adapt the segmentation process to changes in image characteristics caused by variable environmental conditions. The genetic algorithm efficiently searched the hyperspace of segmentation parameter combinations to determine the parameter set which maximized the segmentation quality criteria. A summary of the experimental results that demonstrated the ability to perform adaptive image segmentation and to learn from experience using a collection of outdoor color imagery is presented.

VLSI Design and Implementation of a Real-Time Image Segmentation Processor

Image segmentation was a crucial part of machine vision applications. Presented is a system that performed real-time segmentation of images that used a real-time segmentation VLSI chip that was based on a gradient relaxation algorithm and was designed using the Path Programmable Logic design methodology developed at the University of Utah. The system design considerations, system specifications, and an input/output format for the chip are discussed. The actual design of the chip is given that used pipeline methodology to achieve real-time performance with a compact VLSI layout. The implementation of the segmentation system is presented and the segmentation chip and the overall system are evaluated with regard to real-time performance and segmentation results.

Model-Based Segmentation of FLIR Images

The use of gray scale intensities together with the edge information is presented in a forward-looking infrared (FLIR) image to obtain a precise and accurate segmentation of a target. A model of FLIR images based on gray scale and edge information was incorporated in a gradient relaxation technique which explicitly maximized a criterion function based on the inconsistency and ambiguity of classification of pixels with respect to their neighbors. Four variations of the basic technique were considered which provided automatic selection of thresholds to segment FLIR images. A comparison of these methods is discussed and several examples of segmentation of ship images are given.

Segmentation of Natural Scenes

A simple and computationally efficient approach to image segmentation via recursive region splitting and merging is presented. Unlike other techniques the criterion for splitting was based on a generalization of a two-class gradient relaxation method and merging used a test for mean gray level equivalency for adjacent regions. The technique is illustrated by providing results for both synthetic and natural scenes.

Model-Based Segmentation of FLIR Images

In the automatic recognition of tactical targets in FLIR images, it was desired to obtain an accurate and precise representation of the boundary of the targets. It was very important since the features used in the classification of the target were normally based on the shape and gray scale of the segmented target and therefore the performance of a statistical or a structural classifier critically depended on the results of segmentation. Generally, only the gray scale of the image was used to extract the target from the background and the segmentation thus obtained normally depended upon several parameters of the technique used. It was possible to obtain better segmentation by using other sources of information present in the image such as contextual cues, temporal cues, gradient, a priori information, etc. We considered specifically the use of gray scale together with the edge information present in the image to obtain more precise segmentation of the target than obtained by using gray scale or edge information alone. A model of FLIR images based on gray scale and edge information was incorporated in a gradient relaxation technique which explicitly maximized a criterion function based on the inconsistency and ambiguity of classification of pixels with respect to its neighbors. Four variations of the basic relaxation technique were considered which provide automatic selection of threshold to segment FLIR images.

Segmentation of Images Using a Relaxation Technique

An approach to image segmentation via recursive region splitting is presented. The kernel of the proposed segmentation was based on the two class relaxation technique. An evaluation of this relaxation algorithm was made with respect to the signal to noise ratio, region size, and contrast of the objects present in the image. This established the validity of the two class segmentation technique for segmenting the objects of interest in a multi-class image, when applied on a local basis and recursive manner. The segmentation was analyzed, and its performance on a natural scene is presented.

Segmentation of Images Having Unimodal Distribution

A gradient relaxation method based on maximizing a criterion function was studied and compared to the nonlinear probabilistic relaxation method for the purpose of segmentation of images having unimodal distributions. Although both methods provided comparable segmentation results, the gradient method had the additional advantage of providing control over the relaxation process by choosing three parameters which could be tuned to obtain the desired segmentation results at a faster rate.

Segmentation of Images Using a Gradient Relaxation Technique

A gradient relaxation method based on maximizing a criterion function is presented for the purpose of segmentation of images having almost unimodal distributions. The method provided control over the relaxation process by choosing three parameters which could be tuned to obtain the desired segmentation results.