Chitsobhuk, Orachat

Preferred name

Chitsobhuk, Orachat

Main Affiliation

School of Engineering

orachat.ch@kmitl.ac.th

Now showing 1 - 10 of 15

Some of the metrics are blocked by your
consent settings
An improvement of PDLZW implementation with a modified WSC updating technique on FPGA
(2009-12-01)
Vichitkraivin, Perapong
;
Chitsobhuk, Orachat
In this paper, an improvement of PDLZW implementation with a new dictionary updating technique is proposed. A unique dictionary is partitioned into hierarchical variable word-width dictionaries. This allows us to search through dictionaries in parallel. Moreover, the barrel shifter is adopted for loading a new input string into the shift register in order to achieve a faster speed. However, the original PDLZW uses a simple FIFO update strategy, which is not efficient. Therefore, a new window based updating technique is implemented to better classify the difference in how often each particular address in the window is referred. The freezing policy is applied to the address most often referred, which would not be updated until all the other addresses in the window have the same priority. This guarantees that the more often referred addresses would not be updated until their time comes. This updating policy leads to an improvement on the compression efficiency of the proposed algorithm while still keep the architecture low complexity and easy to implement.
Some of the metrics are blocked by your
consent settings
Color descriptor for image retrieval in wavelet domain
(2006-01-01)
Utenpattanant, Ariya
;
Chitsobhuk, Orachat
;
Khawne, Amnach
This paper presents an approach to manage a large database using a compact color descriptor and a statistical method for accurately pruning the database. A compact color descriptor adopted in the proposed content-based image retrieval system is 63-bit binary Haar color histogram, which is very compact and can be effectively used for fast image search. In addition to fast searching using this compact descriptor, we further improve retrieval time by applying pruning technique, which looks for the candidate images similar to the query image from the database and ignore the rest that are not likely to the query image. The descriptors of the candidate images are then matched with that of the query. The most similar images will be retrieved and ordered according to their distance to the query. The proposed retrieval system can efficiently retrieve the most similar images from the database while can help reducing the retrieval time and the storage space.
Some of the metrics are blocked by your
consent settings
Image registration using hough transform, phase correlation and best-first search algorithm
(2007-12-01)
Chunhavittayatera, Siwaphon
;
Chitsobhuk, Orachat
This paper presents image registration using a combination of Hough transform, phase correlation and best-first search algorithm to estimate rotation and translation parameters. These parameters are used to register input images and create a seamless representation of the registered image. In the first step, the translation of angles are pre-calculated based on ID phase correlation in Hough space and used as candidate angles. Then, best-first search algorithm is applied to obtain the best translation of angle from candidate angles, which is used to de-rotated the input images. In the second step, the translation in x-y axis is computed using 2D phase correlation. Finally, the input images are registered using the estimated translation parameter. The experimental results using various image details and sizes show the accuracy of the proposed technique to detect the translation parameters. The best-first search algorithm can help to increase the precision of rotation parameter after Hough transform and phase correlation while require less amount of processing time compared to full search algorithm.
Some of the metrics are blocked by your
consent settings
Multi-pipeline architecture for face recognition on FPGA
(2009-11-18)
Visakhasart, Sathaporn
;
Chitsobhuk, Orachat
In this paper, a new multi-pipeline architecture is proposed for face recognition system on FPGA. The proposed structure consists of four main units: Multi-Pipeline Control Unit (MPCU), Process Element Unit (PEU), Region Summing Unit (RSU), and Recognition Indexing Unit (RIU). Four recognition techniques: Principal Component Analysis (PCA), Modular PCA (MPCA), Weight MPCA (WMPCA), and Wavelet based techniques are adopted to evaluate the efficiency of the proposed architecture using several standard face databases. The experimental results show that the proposed architecture helps minimizing processing time through its multi-pipeline processes while still maintains high recognition rate. Moreover, the design has encouraged the reduction in hardware resources by utilizing the proposed reusable modules. © 2009 IEEE.
Some of the metrics are blocked by your
consent settings
Image retrieval using connected color region and moment invariants
(2007-12-01)
Dongphontong, Daungkamol
;
Chitsobhuk, Orachat
In this paper, combined descriptors of Connected Color Region (CCR) and moment invariants are proposed as color features for a Content-Based Image Retrieval system. CCR provides the spatial information and maximum co-occurrence color while moment invariants help to better distinguish different distribution of colors in the image. From the experiments, the retrieval results using only CCR descriptors depend on CCR block size. The smaller the size of the block, the higher the retrieval performance. However, if the block size is small, it requires longer retrieval time. Therefore, in this paper, color moment is introduced as additional feature to CCR descriptors to help compromising between retrieval performance and time. The retrieval results using both CCR with large block size and moment descriptors are comparable to those of using only CCR with small block size while require less amount of retrieval time.
Some of the metrics are blocked by your
consent settings
Ultrasonic echo image adaptive watermarking using the just-noticeable difference estimation
(2009-12-01)
Khawne, Amnach
;
Hamamoto, Kazuhiko
;
Chitsobhuk, Orachat
Most of the image watermarking methods, using the properties of the human visual system (HVS), have been proposed in literature. The component of the visual threshold is usually related to either the spatial contrast sensitivity function (CSF) or the visual masking. Especially on the contrast masking, most methods have not mention to the effect near to the edge region. Since the HVS is sensitive what happens on the edge area. This paper proposes ultrasound image watermarking using the visual threshold corresponding to the HVS in which the coefficients in a DCT-block have been classified based on the texture, edge, and plain area. This classification method enables not only useful for imperceptibility when the watermark is insert into an image but also achievable a robustness of watermark detection. A comparison of the proposed method with other methods has been carried out which shown that the proposed method robusts to blockwise memoryless manipulations, and also robust against noise addition.
Some of the metrics are blocked by your
consent settings
An FFT-Based technique and best-first search for image registration
(2008-12-01)
Samritjiarapon, Olan
;
Chitsobhuk, Orachat
The image registration is a fundamental task in image processing used to match two or more pictures taken, for example, at different time, from different sensors, or from different viewpoints. One of the major challenges related to image registration is the estimation of large motion, when input images contain small overlapped area. Common image registration using the search algorithm can accurately finds large motion but requires high computation cost due to large search space. Fourier-based technique is an alternative approach since it can rapidly achieve the registration results through its FFT algorithm. However, only Fourier-based technique cannot produce the correct results in the case of large translation. Thus, this paper presents a Fourier-based technique cooperated with best-first search algorithm to analyze the correct translation between two input images. The Fourier-based technique is used to estimate the candidate translations to decrease searching space while best-first search algorithm is used to further search for the correct translation. The proposed technique can estimate large translations, scalings, and rotations in images by an extension of well-known phase correlation technique. The experimental results using various image details show the accuracy of the proposed technique to detect large translations compared to the other techniques in frequency domain. © 2008 IEEE.
Some of the metrics are blocked by your
consent settings
Image retrieval using haar color descriptor incorporating with pruning techniques
(2007-07-03)
Utenpattanant, Ariya
;
Chitsobhuk, Orachat
This paper presents a content-based image retrieval system using a compact color descriptor (63-bit binary Haar color descriptor) incorporating with pruning techniques. The objective of pruning is to look for the candidate images similar to the query image from the database and ignore the rest that are not likely to the query image. Several statistical pruning criteria are experimented in order to achieve faster retrieval time while preserve good retrieval performance. After the pruning process, the descriptors of the candidate images are then matched with that of the query. The most similar images will be retrieved and ordered according to their distance to the query. The proposed technique is evaluated based on different sizes of the database. The experimental results show good retrieval performance over a variety of image collections.
Some of the metrics are blocked by your
consent settings
Image segmentation based on hierarchical mapping
(2004-12-01)
Junda, Apimun
;
Chitsobhuk, Orachat
In this paper, an efficient image segmentation technique is presented, which combines an image segmentation algorithm with pyramidal approach to form a scale space representation. First, the coarsest level of pyramidal image is quantized to coarse color space. Then, the segmentation is achieved using JSEG algorithm followed by region merging for further refinement. Finally, hierarchical mapping is performed to determine region boundaries in a coarse-to-fine manner using combined global and local features until the final segmentation is accomplished. The multi-resolution approach of the proposed algorithm not only offers a significant reduction in computational cost, but also helps reducing the over-segmentation problem of traditional region growing and watershed techniques. The experimental results show good segmentation performance over a variety of images, also great reduction in the amount of processing time. © 2004 IEEE.
Some of the metrics are blocked by your
consent settings
Performance improvement for arithmetic coder using parallel symbol encoding in JPEG2000
(2008-10-06)
Noikaew, Nopphol
;
Chitsobhuk, Orachat
JPEG2000 image compression standard is designed to support a wide rage of applications. However, its use is restricted since its implementation requires high hardware cost. Bit Plane Coder (BPC) and arithmetic coder are the main resource intensive components, which indicate the overall throughput of JPEG2000 encoder. Usually, the regular operation of the arithmetic coder is sequential, which can process only one symbol at a time. However, the bit-plane coding can generate more than one symbol per clock cycle. Consequently, the coding speed will be limited and bottlenecked at the interface between the output of the bit-plane coding and the input of the arithmetic coder. Therefore, in this paper, the parallel symbol encoding is proposed to improve the performance of the JPEG2000 arithmetic coder. The encoding performance is enhanced through the use of the proposed prediction process of the upper bound value, the lower bound and the index value. Since the prediction processes of the upper bound and lower bound values are quite similar, they can be done simultaneously with the similar prediction condition. As a result, the proposed arithmetic coder architecture can process dual symbols per clock cycle at the speed greater than 100 MHz with the throughput greater than 60 Msymbols/sec. © 2008 IEEE.

Chitsobhuk, Orachat

Filters

Settings

Sort By

Results per page