27
Spatiotemporal Saliency Detection and Its Applications in Static and Dynamic Scenes IEEE TCSVT 2011 Wonjun Kim Chanho Jung Changick Kim

Spatiotemporal Saliency Detection and Its Applications in Static and Dynamic Scenes

  • Upload
    debra

  • View
    84

  • Download
    0

Embed Size (px)

DESCRIPTION

Spatiotemporal Saliency Detection and Its Applications in Static and Dynamic Scenes. IEEE TCSVT 2011 Wonjun Kim Chanho Jung Changick Kim. Outline. Introduction Proposed Method Experiment Result Application Conclusion. Introduction. Problem occurs when background is highly textured. - PowerPoint PPT Presentation

Citation preview

Page 1: Spatiotemporal Saliency Detection and Its Applications in Static and Dynamic Scenes

Spatiotemporal Saliency Detection and Its Applications in Static and Dynamic Scenes

IEEE TCSVT 2011Wonjun Kim

Chanho JungChangick Kim

Page 2: Spatiotemporal Saliency Detection and Its Applications in Static and Dynamic Scenes

OutlineIntroductionProposed MethodExperiment ResultApplicationConclusion

Page 3: Spatiotemporal Saliency Detection and Its Applications in Static and Dynamic Scenes

IntroductionProblem occurs when background is highly textured

Page 4: Spatiotemporal Saliency Detection and Its Applications in Static and Dynamic Scenes

Proposed Method

Feature RepresentationEdge orientation histogram (EOH)Color orientation histogram (COH)Temporal Feature

Self-ordinal MeasureSaliency MapScale-invariant Saliency Map

Page 5: Spatiotemporal Saliency Detection and Its Applications in Static and Dynamic Scenes

Edge Orientation Histogram (EOH)

1. Compute the edge orientation of every pixel in the local region center at the pixel

2. Quantized into K angle in the range of [,]3. Compute the histogram of edge orientation

m(x,y,n):edge magnitude(x,y,n):quantized orientation

𝑖

local region

Page 6: Spatiotemporal Saliency Detection and Its Applications in Static and Dynamic Scenes

Color Orientation Histogram (COH)

1. Quantize the angle in HSV color space in the range of [,] into H angles

2. Compute the histogram of color orientation

s(x,y,n):saturation value(x,y,n):quantized hue value

Page 7: Spatiotemporal Saliency Detection and Its Applications in Static and Dynamic Scenes

Temporal FeatureCompute the intensity differences between frames

Feature at the pixel of frame

P :total number of pixels in local regionj :index of those pixels in P :user-defined latency

Page 8: Spatiotemporal Saliency Detection and Its Applications in Static and Dynamic Scenes

Self-ordinal MeasureDefine a 1(K+1) rank matrix by ordering the

elements of EOH(COH) ex:

Page 9: Spatiotemporal Saliency Detection and Its Applications in Static and Dynamic Scenes

Self-ordinal Measure

Page 10: Spatiotemporal Saliency Detection and Its Applications in Static and Dynamic Scenes

Saliency Map of Edge and ColorCompute the distance from the rank matrix of

center region to surrounding regions

Saliency Map of Edge Saliency Map of Color

N :total number of local regions in a center-surround window

, :maximum distance between two rank matrices

Page 11: Spatiotemporal Saliency Detection and Its Applications in Static and Dynamic Scenes

Spatial Saliency MapCombine the edge and color saliency

Page 12: Spatiotemporal Saliency Detection and Its Applications in Static and Dynamic Scenes

Combining with Temporal SaliencyCompute the SAD of temporal gradients between

center and the surrounding regions

Combine the spatial and temporal saliency

Page 13: Spatiotemporal Saliency Detection and Its Applications in Static and Dynamic Scenes

Scale-invariant Saliency MapCombine 3 different scales of saliency Map

(3232, 6464, 128128)

3232 1281286464

Page 14: Spatiotemporal Saliency Detection and Its Applications in Static and Dynamic Scenes

Algorithm

Page 15: Spatiotemporal Saliency Detection and Its Applications in Static and Dynamic Scenes

Experiment Result

Static ImagesVideo Sequences

Page 16: Spatiotemporal Saliency Detection and Its Applications in Static and Dynamic Scenes

Experiment ResultStatic Image

Local region = 55center-surround window = 77K = 8, H= 6 = 40, = 24

Video Sequence = 49Speed: 23ms per frame (43 fps)

Page 17: Spatiotemporal Saliency Detection and Its Applications in Static and Dynamic Scenes

Static Images

Page 18: Spatiotemporal Saliency Detection and Its Applications in Static and Dynamic Scenes

Static Images

Page 19: Spatiotemporal Saliency Detection and Its Applications in Static and Dynamic Scenes

Video Sequences

Page 20: Spatiotemporal Saliency Detection and Its Applications in Static and Dynamic Scenes

Video Sequences

Page 21: Spatiotemporal Saliency Detection and Its Applications in Static and Dynamic Scenes

Application

Image RetargetingMoving Object Extraction

Page 22: Spatiotemporal Saliency Detection and Its Applications in Static and Dynamic Scenes

Image Retargeting

Page 23: Spatiotemporal Saliency Detection and Its Applications in Static and Dynamic Scenes

Image Retargeting

Page 24: Spatiotemporal Saliency Detection and Its Applications in Static and Dynamic Scenes

Moving Object DetectionG:the set of salient pixels in the ground truth imageP:salient pixels in the binarized object mapCard(A):the size of the set A

Page 25: Spatiotemporal Saliency Detection and Its Applications in Static and Dynamic Scenes

Moving Object Detection

Page 26: Spatiotemporal Saliency Detection and Its Applications in Static and Dynamic Scenes
Page 27: Spatiotemporal Saliency Detection and Its Applications in Static and Dynamic Scenes

ConclusionOrdinal signature can tolerate more local feature

distribution than sample values.The proposed scheme performs in real-time and

can be extended in both static and dynamic scenes.