Gwangwon Kang

Efficient Image Retrieval Using Adaptive Segmentation of HSV Color Space

This paper presents an efficient image retrieval system using adaptive segmentation of hue, saturation and value (HSV) color space. We classify the image into n number of areas based on different selected ranges of hue and value, then each area is partitioned into m number of segments based on the number of pixels it contains, the area which has more pixels will be partitioned into more segments and the area which has less pixels will be partitioned into less number of segments. This is used as a feature vector. Retrieval system outputs the image with a high matching factor. A small demonstration system has been tested and shows superior performance compared with the simple color based retrieval systems.

Sum of Values of Local Histograms for Image Retrieval

CBIR makes a wide use of histogram based methods for image indexing. Histograms describe the global intensity distribution of images. They are very easy to compute and are insensitive to small changes in object translations and rotations. However, they are not robust to large appearance changes, and they might give similar results for different kinds of images if the distributions of colors are same in the images. Our research focuses mainly on the image bins (histogram value divisions by frequency) separation technique followed by calculating the sum of values, and using them as image local features. At first, the histogram is first calculated for an image. After that, it is subdivided into sixteen equal bins and the sum of local values is calculated and stored. We have tested the proposed algorithm on a large database of images.

Shape from Focus through Laplacian Using 3D Window

One of the fundamental objectives of computer vision is to reconstruct a three-dimensional (3D) structure of objects from two-dimensional (2D) images. The basic idea of image focus is that objects at different distances from a lens are focused at different distances. Shape from Focus (SFF) is the problem of reconstructing the depth of the scene changing actively the optics of the camera until the point of interest is in focus. The point in focus gives information about its depth through the thin lens Gaussian law. An effective focus measure operator should be a high-pass filter. Usually, the variation of frequency components are not enough that focus measure could be computed pixel-wise, therefore, sum of pixels in small 2D windows are used for detecting the high frequency components. In this paper, we propose to use 3D windows instead of 2D windows for detecting the high frequency components in the images. The proposed algorithm using 3D window gives better depth map than the previous algorithms using 2D windows.

Features Defined by Median Filtering on RGB Segments for Image Retrieval

The performance of image retrieval using median filtering on RGB color information was analyzed in this paper in order to design a more effective algorithm for extracting features from color images for image retrieval. We propose an image retrieval technique, which uses features obtained by indexing color information. The method uses size order and quantization after intermediate values are extracted for each RGB image and partitioned into regular sized blocks. Small feature table based on color image features are proposed in this paper, because even an effective feature extraction algorithm requires a large amount of storage space and calculation. Matches were obtained by comparing normalized values of features that were organized into a table using the proposed algorithm, for the input image and existing images and reorganizing into a table that can use correlogram.

Image Indexing using Spatial Multi-Resolution Color Correlogram

Color correlograms are efficiently used for image indexing in content-based image retrieval. Color correlogram extracts not only the color distribution of pixels in images like color histogram, but also extracts the spatial information of pixels in the images. The characteristic of the color correlogram to take into account the spatial information as well as the distribution of color pixels greatly attracts the researcher for content based image retrieval. Even though, a single correlogram is not enough for efficient and robust image retrieval system. In this paper, we propose the use of color correlogram on multiresolution images. The multiresolution color correlogram gives much better retrieval efficiency, but with higher computations. The multiresolution images are generated using the median filters.

A novel algorithm for identification of body parts in medical images

In this paper, we introduce an algorithm based on energy information obtained from Wavelet Transform for classification of medical images according to imaging modalities and body parts. Various medical image retrieval systems are available today that classify images according to imaging modalities, orientations, body parts or diseases. Generally these are limited to either some specific body part or some specific disease. Further, almost all of them deal with the DICOM imaging format. Our technique, on the other hand, can be applied to any of the imaging formats. The results are shown for JPEG images in addition to DICOM imaging format. We have used two types of wavelets and we have shown that energy obtained in either case is quite distinct for each of the body part.

Defining a Set of Features Using Histogram Analysis for Content Based Image Retrieval

A new set of features are proposed for Content Based Image Retrieval (CBIR) in this paper. The selection of the features is based on histogram analysis. Standard histograms, because of their efficiency and insensitivity to small changes, are widely used for content based image retrieval. But the main disadvantage of histograms is that many images of different appearances can have similar histograms because histograms provide coarse characterization of an image. Hence we further refine the histogram using the histogram refinement method. We split the pixels in a given bucket into several classes just like histogram refinement method. The classes are all related to colors and are based on color coherence vectors. After the calculation of clusters using histogram refinement method, inherent features of each of the cluster is calculated. These inherent features include size, mean, variance, major axis length, minor axis length and angle between x-axis and major axis of ellipse for various clusters.

Object Retrieval Approach with Invariant Features Based on Corner Shapes

This paper presents a new technique for corner shape based object retrieval from a database. The proposed feature matrix consists of values obtained through a neighborhood operation of detected corners. This result in a significant small size feature matrix compared to the algorithms using color features and thus is computationally very efficient. The corners have been extracted by finding the intersections of the detected lines found using Hough transform. As the affine transformations preserve the co-linearity of points on a line and their intersection properties, the resulting corner features for image retrieval are robust to affine transformations. Furthermore, the corner features are invariant to noise. It is considered that the proposed algorithm will produce good results in combination with other algorithms in a way of incremental verification for similarity.

Design of a Circularly Polarized Z-Slot Antenna with Isotropic Pattern for the UHF RFID Reader of WSN

In an antenna for a UHF RFID reader of wireless sensor networks (WSN), receiver sensitivity in sensing multitags from remote distances is an important performance index. This study designed a dual structured Z-slot antenna with optimized receiver sensitivity to enhance the sensitivity to a circularly polarized antenna with an isotropic pattern for a UHF RFID. Through analysis of performance in the designed antenna, the following was verified: return loss ( S 11 ) was about −62.21 dB at 925.25 MHz, antenna gain was 7.36 dBi, and Δ P r , isotropic gain deviation, was 1.3 dB. Impedance matching was about 50.069 Ω at 925.25 MHz and VSWR was from 1.001 to 1.028. Through this research it was discovered that this can be applied to the design of all RFID readers of WSN. Based on the above results, it is suggested that a circularly polarized Z-slot antenna which can enhance receiver sensitivity over a wide range can be widely applied to UHF RFID readers of WSN.

Image Feature Vector Construction Using Interest Point Based Regions

The paper presents a new approach for content based retrieval of images. The algorithm uses information sampled from around detected corner points in the image. A corner detection approach based on line intersections has been employed using Hough transform for line detection and then finding intersecting, near intersecting or complex shaped corners. As the affine transformations preserve the co-linearity of points on a line and their intersection properties, the corner points obtained as such retain the much desired property of repeatability and hence ensure the similar pixel samples under various transformations and are robust to noise. K-means clustering algorithm is used to assign class labels to the extracted sample mean and variance of the corner regions from a random selection of training images and used for learning a Gaussian Byes classifier to classify whole training image database. Histogram of the class members in an image has been used as a feature vector. The retrieval performance and behavior of the algorithm has been tested using four different similarity measures.