Wavelet-Based Multiresolution Matching for

Content-Based Image Retrieval

Presented by

Tienwei TsaiDepartment of Computer Science and Engineering

Tatung University2005/12/10

Outline

1. Introduction 1. Introduction

2. Related Works2. Related Works

3. Proposed Image Retrieval System3. Proposed Image Retrieval System

4. Experimental Results4. Experimental Results

5. Conclusions5. Conclusions

1. Introduction

• Two approaches for image retrieval: – query-by-text (QBT): annotation-based image

retrieval (ABIR)– query-by-example (QBE): content-based

image retrieval (CBIR)

• Standard CBIR techniques can find the images exactly matching the user query only.

• In QBE, the retrieval of images basically has been done via the similarity between the query image and all candidates on the image database. – Euclidean distance

• Transform type feature extraction techniques– Wavelet, Walsh, Fourier, 2-D moment, DCT, and

Karhunen-Loeve.

• In our approach, the wavelet transform is used to extract low-level texture features.

2. Related Works

• Content-based image retrieval is a technology to search for similar images to a query based only on the image pixel representation.– However, the query based on pixel information is quite time-

consuming– Therefore, how to choose a suitable color space and reduce the

data to be computed is a critical problem in image retrieval.

• Some of the systems employ color histograms. – The histogram measures are only dependent on summations of

identical pixel values and do not incorporate orientation and position.

• Therefore, we propose an image retrieval scheme to retrieve images from their transform domain, which tries to reduce data and still retains their local information.

• In this paper, we focus on the QbE approach. The user gives an example image similar to the one he/she is looking for.

• Finally, the images in the database with the smallest distance to the query image will be given, ranking according to their similarity. – We can define the QbE problem as follows. Given a

query image Q and a database of images X1, X2,…, Xn, find the image Xi closest to Q. The closeness is to be computed using a distance measuring function D(Q, Xn).

3. The Proposed Image Retrieval System

Figure 1. The proposed system architecture.

Feature Extraction

• Features are functions of the measurements performed on a class of objects (or patterns) that enable that class to be distinguished from other classes in the same general category.

• Color Space TransformationRGB (Red, Green, and Blue) ->

YUV (Luminance and Chroma channels)

YUV color space

• YUV is based on the CIE Y primary, and also chrominance.– The Y primary was specifically designed to follow the

luminous efficiency function of human eyes. – Chrominance is the difference between a color and a

reference white at the same luminance.

• The following equations are used to convert from RGB to YUV spaces:

– Y(x, y) = 0.299 R(x, y) + 0.587 G(x, y) + 0.114 B(x, y),

– U(x, y) = 0.492 (B(x, y) - Y(x, y)), and

– V(x, y) = 0.877 (R(x, y) - Y(x, y)).

Wavelet Transform

• Mallat' s pyramid algorithm

Figure 2.

Figure 3.

Distance Measurement

• In our experimental system, we define a measure called the sum of squared differences (SSD) to indicate the degree of distance (or dissimilarity).

• The distance between Q and Xn under the LL(k) subband can be defined as

m n

kx

kqnLL

nmLLnmLLXQDn

k

2)()( ),(),(),()(

• Based on the observation that subimages LL(1), LL(2), …, LL(k) correspond to different resolution levels of the original image, the retrieving accuracy may be improved to consider these subimages at the same time.

• Therefore, the distance between Q and Xn can be modified as the weighted combination of these subimages:

where wk is the weight of the distance under the kth resoluiton level.

m

nLLkn XQDwXQD k ),(),( )(

4. Experimental Results

• 1000 images downloaded from the WBIIS database are used to demonstrate the effectiveness of our system.

• The user can query by an external image or an image from the database.

• To evaluate the retrieval efficiency of the proposed method, we use the performance measure, the precision rate, as follows:

where Rr is the number of relevant retrieved items, and Tr is the number of all retrieved items.

r

r

T

Rrateprecision

Figure 4. Retrieved results via the comparison of the original RGB images.

Figure 5. Retrieved results based on the Y-component images.

Figure 6. Retrieved results based on the Y-component LL(1) subimages.

Figure 7. Retrieved results based on the Y-component LL(2) subimages.

Figure 8. Retrieved results based on the Y-component LL(3) subimages.

Figure 9. Retrieved results based on the Y-component LL(4) subimages.

Figure 10. Retrieved results based on the Y-component LL(5) subimages.

Figure 11. Retrieved results based on the Y-component LL(6) subimages.

Figure 12. Retrieved results based on the combination of the Y-component LL(2) and LL(3) subimages.

Figure 13. Retrieved results based on the combination of the Y-component LL(3), LL(4), and LL(5) subimages.

5. Conclusions

• In this paper, we propose a CBIR method based on the color space transformation and the wavelet transform.

• We find that through the combination of the features with different resolution level, we can not only obtain the better precision rate but also the good reduction rate.

Future Works

• Since only preliminary experiment has been made to test our approach, a lot of works should be done to improve this system: – Since several features may be used simultaneously, it

is necessary to develop a scheme that can integrate the similarity scores resulting from the matching processes.

– A long-term aim is combining the semantic annotations and low-level features to improve the retrieval performance. That is, the retrieved images should be somehow related to the objects contained in the scenes.

Thank You !!!