Chitwong, Sakreya

Preferred name

Chitwong, Sakreya

Alternative Name

Chitwong, S.

Main Affiliation

School of Engineering

sakreya.ch@kmitl.ac.th

Now showing 1 - 9 of 9

Some of the metrics are blocked by your
consent settings
A fast intensity-hue-saturation fusion approach via principal component analysis for ikonos imagery
(2008-12-01)
Minhayenud, S.
;
Chitwong, S.
;
Cheevasuvit, F.
To enhance spatial information of low resolution multi-spectral (RGB) image, the intensity-hue-saturation (IHS) approach is perfectly used to fuse the low resolution RGB image and the high resolution panchromatic (Pan) image by replacing intensity component with the high-resolution Pan image. Disadvantage of the mentioned approach is that color of a fused image is changed because the saturation component is changed or spectral of the low resolution RGB image and the high resolution Pan image is different, that is, spectral information of the fused RGB image is distorted. This problem is important for applying the fused image for classification. To solve this problem, in this paper, we employ the principal component analysis (PCA) transformation to extract information from the low resolution RGB image. In procedure of fusion method, the first principal component is used to adjust brightness of the high resolution Pan image. The intensity component from IHS transformation is replaced by the adjusted brightness high-resolution Pan image. The experimental results by using IKONOS imagery show that the proposed approach is better performance than the original IHS methods by improving spectral distortion and still correlating to the Pan image.
Some of the metrics are blocked by your
consent settings
Speckle noise reduction using adaptive singular value decomposition in logarithmic domain
(2005-12-01)
Chitwong, S.
;
Thongsila, A.
;
Intajag, S.
;
Nilas, P.
;
Cheevasuvit, F.
This paper presents applying the singular value decomposition to reduce speckle noise. Generally, it is used to filter the additive Gaussian noise with zero mean and any variance. Since speckle noise is in multiplicative model, to transform multiplicative model into additive model, we then employ logarithmic transformation. In this paper, speckle noise is generally modeled as Gamma distribution function corresponding with speckle noise of synthetic aperture radar (SAR) imagery applied. All singular value decomposition based filtering processing is in logarithmic domain. Threshold value to determine the effective rank and orders of matrix are adapted as homogeneity analysis. The orders of matrix are consisted of 16 by 16, 8 by 8 and 4 by 4. Normally, the results of the singular value decomposition based filtering after that the filtered matrix is transformed into spatial domain by exponential function is in block-fashion, then blocking effect is occurred. To smooth, the filtered matrix is performed as average filtering by using a number of pixels of 4 by 4 pixels around center of one. Experiments are tested using both simulated image and real image. Signal to noise ratio and equivalent number of looks are employed to evaluate the performance of our method. Our results are good enough when compared with the recent results at which such method is more complex. © 2005 by the American Society for Photogrammetry and Remote Sensing.
Some of the metrics are blocked by your
consent settings
Enhancement of color image obtained from PCA-FCM technique using local area histogram equalization
(2002-12-01)
Chitwong, S.
;
Boonmee, T.
;
Cheevasuvit, F.
This paper presents local area enhancement of the segmented color image obtained from the multi-spectral image clustering by using FCM (fuzzy c-means). In case, the multi-spectral images, which have the number of bands more than that of 3, must decrease the data volume to remain the number of bands of 3 in order to correspond with the meaning of red, green, and blue images. PCA (Principal Components Analysis) is then used to transform original multi-spectral images into PCA images. The first three components having information more than that of original images of 95% is assigned as red, green, and blue images, namely RGB color image. FCM clustering apply to RGB color image, separately. This method is called the PCA-FCM technique being the multi-spectral image clustering. By applying such technique, the result images consisted of red, green, and blue images separately are the segmented images. By histogram equalization algorithm, the result of local area enhancement based on a number of clusters as the segmented image can solve effect of intensity saturation from global area enhancement and the perceptibility of color image is clearly improved.
Some of the metrics are blocked by your
consent settings
Segmentation on edge preserving smoothing image based on graph theory
(2000-12-01)
Chitwong, S.
;
Cheevasuvit, F.
;
Dejhan, K.
;
Mitatha, S.
;
Nokyoo, C.
The presented of noise in an image will cause many undesired small holes in the segmented image. This effect causes an important efficiency decreasing for classifying and describing objects. To remove the embed noise while the edges of the image still preserving, the edge preserving smoothing process must be applied. The smoothing process replaces the pixel intensity of the considered pixel by the average intensity of the most homogeneous mask among the proposed masks. The proposed masks can be preserved the thin region even its width is less than 3 pixels then, the smoothed image will be segmented by graph theory in order to obtain the higher accurate region's boundaries. In the mean time of segmentation process, the homogeneous threshold value has been applied to ensure that the maximum different gray value of each segmented region is controlled.
Some of the metrics are blocked by your
consent settings
Contrast enhancement of satellite image based on adaptive unsharp masking using wavelet transform
(2006-12-01)
Chitwong, S.
;
Phahonyothing, S.
;
Nilas, P.
;
Cheevasuvit, F.
This paper concerns with a method for unsharp masking for contrast enhancement of satellite image. We employ the nature of wavelet transform that separates the original image into low and high frequency sub-band images as low and high pass filter. Particularly, a number of high frequency sub-band images consist of horizontal, vertical, and diagonal coefficients that contain detail of information. Taking inverse wavelet transform of each sub-band image separately except low frequency one, we have each of high frequency information in horizontal, vertical, and diagonal image. All of them are scaled by the scaling factor in each one separately. Adaptive algorithm is implemented to results the suitable scaling factor to obtain the enhanced image corresponding with the given criterion based on variance of each area smooth and detail area. Experimental results show that our method performs well to high enhance in detail area and low in smooth area.
Some of the metrics are blocked by your
consent settings
Fusion of principal component of multispectral bands with PAN band using HIS and wavelet transform
(2002-01-01)
Chitwong, S.
;
Cheevasuvit, F.
;
Homthong, J.
Since RGB images derived from multispectral (TM) images will lose some information, in this paper we present the method to solve such problem by using principal component analysis (PCA) which transforms TM images into the principal component images (PCs), while the high resolution PAN data is decomposed by wavelet transform. Thus, RGB images are assigned by the first three principal component images which normally have approximately 95% of the information in the original images. The intensity image from RGB to HIS transformation is replaced by the lower frequency coefficient of wavelet transform of PAN data corresponding to multispectral images. HIS to RGB transformation is then applied. The fused RGB image using our method can obtain more details.
Some of the metrics are blocked by your
consent settings
Local area histogram equalization based multispectral image enhancement from clustering using competitive hopfield neural network
(2003-10-01)
Chitwong, S.
;
Boonmee, T.
;
Cheevasuvit, F.
One of important issues for enhancing image based on local area histogram equalization (LHE) is a clustering or segmenting technique. That is, the more the accuracy of separating image into specified classes is needed, the better the performance of enhancement is. As mentioned objective, in this paper, the competitive Hopfield neural network (CHNN) is then proposed for clustering to the LHE based image enhancement. By using simulated image, standard image and mutispectral image from Landsat 7 satellite, experimental results are shown in both accuracy of clustering and variance of the enhanced image. The criteria for a good enhancement algorithm is that it can give high variance in detail area, low variance in smooth and edge areas. Also comparing the variance of the enhanced image by both LHE and global area histogram equalization (GHE) methods shows that one from LHE outperforms. In addition, the enlarged image from small area is shown clearly by visualization. All results compare with the conventional methods such as fuzzy c-means (FCM).
Some of the metrics are blocked by your
consent settings
Image coding using adaptive vector quantization of wavelet coefficients
(2001-01-01)
Chitwong, S.
;
Cheevasuvit, F.
;
Sinthuvanichsaid, J.
In this paper we propose a subband image compression by using wavelet transform to split original images. Each of subband images is then quantized by an adaptive vector quantization with dynamic bit allocation based on advantage of nature of wavelet coefficients. The energy of each subband image, except the lowest frequency subband image will not be quantized, will be sorted from minimum to maximum. Energy of each subband image is calculated to allocate bits not over the desired bit rate. The accumulation of energy from these subband images will be divided into 4 groups. First two lower energy groups will be encoded with 256 and 16 code vectors for 16 pixels block size in accordance with energy ratio. Others will be encoded with 256 code vectors for 4 and 16 pixels block size. Based on the given bit rate, the total dynamical bit rate of each group is calculated. If the total dynamical bit rate in the group is less or more than the given bit, it will then be adjusted based on the energy of subband image in only the same group. The remaining of energy from higher energy group will be carried to lower. The experiments are shown that the resulting images from the proposed, method can be clearly improved by Peak Signal to Noise Ratio (PSNR) of 36.3016, MSE = 15.2377, 1.03125 bpps.
Some of the metrics are blocked by your
consent settings
Contrast enhancement for mimimum mean brightness error from histogram partitioning
(2009-12-01)
Phanthuna, N.
;
Cheevasuvit, F.
;
Chitwong, S.
This paper presents the image enhancing using a mean separated histogram equalization method. To provide the minimum mean brightness error after the histogram modification. It separates the input image's histogram into n (n=1,2,3,⋯) groups based on input mean before equalizing them independently. The image initially is separated class by calculated threshold level and each class is histogram equalized to entire image, and gets lowest AMBE (AMBE: Absolute Mean Brightness Error). The result found that AMBE gradually reduces when the separation is increased. Therefore, the error threshold is assigned in order to automatically dividing the original histogram for obtaining the desired AMBE. This process will be applied to remote sensing data by treating each region of histogram independently. Also Tenengrad is employed in order to verify the contrast performance. The image performance is considered higher if its Tenengrad value is larger.

Chitwong, Sakreya

Filters

Settings

Sort By

Results per page