36
A Wrapper-Based Approach to Image Segmentation and Classification Michael E. Farmer, Member, I EEE, and Anil K. Jain, Fellow, IEEE

A Wrapper-Based Approach to Image Segmentation and Classification

  • Upload
    adamma

  • View
    35

  • Download
    0

Embed Size (px)

DESCRIPTION

A Wrapper-Based Approach to Image Segmentation and Classification. Michael E. Farmer , Member, IEEE, and Anil K. Jain , Fellow, IEEE. 大綱. Introduction Overview of the approach Experiment: Vision-Base airbag suppression application - PowerPoint PPT Presentation

Citation preview

Page 1: A Wrapper-Based Approach to Image Segmentation  and Classification

A Wrapper-Based Approach to Image Segmentation and Classification

Michael E. Farmer, Member, IEEE, and Anil K. Jain, Fellow, IEEE

Page 2: A Wrapper-Based Approach to Image Segmentation  and Classification

大綱大綱 IntroductionIntroduction Overview of the approachOverview of the approach Experiment: Vision-Base airbag suppression Experiment: Vision-Base airbag suppression

applicationapplication Experimental resultExperimental result

Page 3: A Wrapper-Based Approach to Image Segmentation  and Classification

IntroductionIntroduction

Page 4: A Wrapper-Based Approach to Image Segmentation  and Classification

Traditional processingTraditional processing

The traditional processing flow for image-based pattern recognition consists of image segmentation followed by classification.

Page 5: A Wrapper-Based Approach to Image Segmentation  and Classification

Three limitations of traditional Three limitations of traditional processingprocessing

The object of interest “should be uniform and homogeneous with respect to some characteristic” and “adjacent regions should be differing significantly”

There are few metrics available for evaluating segmentation algorithms

Inability to adapt to real-world changes

Page 6: A Wrapper-Based Approach to Image Segmentation  and Classification

The contributions in this paper

Developing a closed-loop framework for image segmentation to find the best segmentation for a given class of objects by using the shape of the object for classification of the segmented object

Using the probability of correct classification of the object to provide an “objective evaluation of segmented outputs”

The system can adapt to “real-world changes.”

Page 7: A Wrapper-Based Approach to Image Segmentation  and Classification

Overview of the approachOverview of the approach

Page 8: A Wrapper-Based Approach to Image Segmentation  and Classification

Wrapper-Based ApproachWrapper-Based Approach

Wrap the segmentation and the classification together, and use the classifier as the metric for selecting the best segmentation.

Using the classifier to intelligently re-assemble to solve over-segmented problem.

The classification is correct when the minimum distance between the classification of the candidate segmentation and one of the desired pattern classes < T

Page 9: A Wrapper-Based Approach to Image Segmentation  and Classification

Traditional vs Wrapper-BaseTraditional vs Wrapper-Base

Page 10: A Wrapper-Based Approach to Image Segmentation  and Classification

Experiment: Experiment: Vision-Base airbag suppression Vision-Base airbag suppression

applicationapplication

Page 11: A Wrapper-Based Approach to Image Segmentation  and Classification

ProblemProblem

Infant or Adult

Page 12: A Wrapper-Based Approach to Image Segmentation  and Classification

ChallengesChallenges

Nonuniform illumination Poor image contrast Shadows and highlights Occlusions Sensor noise Background clutter

Page 13: A Wrapper-Based Approach to Image Segmentation  and Classification

Variability for the infant classVariability for the infant class

Page 14: A Wrapper-Based Approach to Image Segmentation  and Classification

Variability for the infant classVariability for the infant class

Page 15: A Wrapper-Based Approach to Image Segmentation  and Classification

Proposed approachProposed approach

Page 16: A Wrapper-Based Approach to Image Segmentation  and Classification

Preliminary SegmentationPreliminary Segmentation Reduce the number of blobs that must be processed.

Once the correlation value for each region is determined, an adaptive threshold is applied, and any region that falls below the threshold is considered a part of the foreground.

Page 17: A Wrapper-Based Approach to Image Segmentation  and Classification

Preliminary SegmentationPreliminary Segmentation

Page 18: A Wrapper-Based Approach to Image Segmentation  and Classification

Preliminary SegmentationPreliminary Segmentation

Page 19: A Wrapper-Based Approach to Image Segmentation  and Classification

RegionRegion LabelingLabeling

Using the EM algorithm with a fixed number of components, and then rely on the classification accuracy to determine if more components are required.

Merging the very small blobs by mode filter Merging any regions that are smaller then 20

pixels in size with their larger neighbors

Page 20: A Wrapper-Based Approach to Image Segmentation  and Classification

RegionRegion Labeling ResultsLabeling Results

Page 21: A Wrapper-Based Approach to Image Segmentation  and Classification

RegionRegion Labeling ResultsLabeling Results

Page 22: A Wrapper-Based Approach to Image Segmentation  and Classification

Blob CombinerBlob Combiner

We have framed the blob combiner problem as one of blob selection, where there exists a subset of blobs that will provide the highest classification accuracy for a given pattern class

Forward selection modeForward selection mode Backward selection modeBackward selection mode

Page 23: A Wrapper-Based Approach to Image Segmentation  and Classification

Blob CombinerBlob Combiner( ( plus-L, minus-R algorithm )

Page 24: A Wrapper-Based Approach to Image Segmentation  and Classification

Blob CombinerBlob Combiner ( ( plus-L, minus-R algorithm )

Page 25: A Wrapper-Based Approach to Image Segmentation  and Classification

Feature ExtractionFeature Extraction

Page 26: A Wrapper-Based Approach to Image Segmentation  and Classification

Feature ExtractionFeature Extraction

Page 27: A Wrapper-Based Approach to Image Segmentation  and Classification

Acceleration Methods forAcceleration Methods for Feature Extraction Feature Extraction:

Precompute the moments for each blob Compute the moments using only the local nei

ghborhood of each blob.

Attain over a ten thousand-fold reduction in processing for each moment calculated.

Page 28: A Wrapper-Based Approach to Image Segmentation  and Classification

Classification of Blob CombinationsClassification of Blob Combinations

Using the nearest neighbor classifier to compute classification distance

Feature 1

Fea

ture

2

: class - A points: class - B points: point with unknown class

Circle of 1 - nearest neighborThe point is class B via 1-NNR.

Page 29: A Wrapper-Based Approach to Image Segmentation  and Classification

Proposed approachProposed approach

Page 30: A Wrapper-Based Approach to Image Segmentation  and Classification

Demonstrating

Page 31: A Wrapper-Based Approach to Image Segmentation  and Classification

Demonstrating

Page 32: A Wrapper-Based Approach to Image Segmentation  and Classification

Demonstrating

Page 33: A Wrapper-Based Approach to Image Segmentation  and Classification

EXPERIMENTAL RESULTS

Page 34: A Wrapper-Based Approach to Image Segmentation  and Classification

EXPERIMENTAL RESULTS

Page 35: A Wrapper-Based Approach to Image Segmentation  and Classification

Correct segmentations

Page 36: A Wrapper-Based Approach to Image Segmentation  and Classification

Incorrect segmentation