• Journal of the Institute of Electronics Engineers of Korea SP
• /
• v.42 no.6
• /
• pp.19-26
• /
• 2005

Image interpolation in the wavelet domain usually takes advantage of the probabilistic models for the intrascale statistics and the interscale dependency. In this paper, we adopt the linear model for the absolute values of wavelet coefficients of interpolated image across scale to estimate the variances of extrapolated bands. The proposed algorithm uses randomly generated wavelet coefficients based on the estimated parameters for probabilistic model. Random number generation according to the estimated probabilistic model may induce the 'salt and pepper' noise in subbands. We reduce the noise power by Wiener filtering. We observe that the proposed method generates the histogram of the subband coefficients similar to the that of original image. Experimental results show that our method outperforms the previous wavelet-domain interpolation method as well as the conventional bicubic method.

• Journal of Korea Multimedia Society
• /
• v.10 no.9
• /
• pp.1125-1134
• /
• 2007

In this paper, we propose a new approach for rotation invariant texture classification based on Gabor wavelet. Conventional methods have the low correct classification rate in large texture database. In our proposed method, we define two feature groups which are the global feature vector and the local feature matrix. The feature groups are output of Gabor wavelet filtering. By using the feature groups, we defined an improved discriminant and obtained high classification rates of large texture database in the experiments. From spectrum symmetry of texture images, the number of test times were reduced nearly 50%. Consequently, the correct classification rate is improved with $2.3%{\sim}15.6%$ values in 112 Brodatz texture class, which may vary according to comparison methods.

• Journal of Korea Multimedia Society
• /
• v.11 no.4
• /
• pp.443-453
• /
• 2008

In this paper, the 1-level DWT(Discrete Wavelet Transform) coefficients of a binary logo image are used as the watermark. The watermark should be inserted in the same band which is equivalent to the host image when the watermark is inserted in the wavelet domain. This is the image fusion of the proposed watermarking method. The watermark is inserted in relatively significant coefficients after the insertion area is defined. The more significant coefficients have the important information because they are identified as the edge and major surface in images. The significant coefficients are defined when their absolute value exceeds the threshold. The standard deviation is used as the weight value of watermark insertion in order to strengthen the weight of the watermark insertion according to the value of the coefficients. The proposed watermarking method is an adaptive scheme, and the proposed two detection algorithms can be adaptively used when the watermarked image is distorted by cropping, filtering, or compression.

• Journal of Broadcast Engineering
• /
• v.26 no.6
• /
• pp.725-737
• /
• 2021

Although the performance of cameras is gradually improving now, there are noise in the acquired digital images from the camera, which acts as an obstacle to obtaining high-resolution images. Traditionally, a filtering method has been used for denoising, and a convolutional neural network (CNN), one of the deep learning techniques, has been showing better performance than traditional methods in the field of image denoising, but the details in images could be lost during the learning process. In this paper, we present a CNN for image denoising, which improves image details by learning the details of the image based on wavelet transform. The proposed network uses two subnetworks for detail enhancement and noise extraction. The experiment was conducted through Gaussian noise and real-world noise, we confirmed that our proposed method was able to solve the detail loss problem more effectively than conventional algorithms, and we verified that both objective quality evaluation and subjective quality comparison showed excellent results.

• Proceedings of the IEEK Conference
• /
• 2001.09a
• /
• pp.907-910
• /
• 2001

In this paper, we address a new time delay estimation method using third-order statistics and subband adaptive filtering to improve the accuracy of target detection for acoustic backscattered signals in a noise interference environment. Each reference and primary signals are decorrelated using the multiresolution analysis framework through a M-band discrete wavelet transform(M-DWT). Then noise effect can be reduced. Here, time delays are estimated iteratively in each subband using two different adaptation mechanisms that minimize the mean squared error (MSE) between the references and primary signal. More specifically, third-order cumulants and projection cross-correlation(PCC) criterion are utilized to achieve an effective SNR improvement for the time delay estimation.

• Proceedings of the IEEK Conference
• /
• 2003.07e
• /
• pp.1815-1818
• /
• 2003

Digital watermarking is technique, which owner's information is inserted in digital image, with intention to protecting a copyright of digital image. In watermarking for copyright and authentication, watermark shouldn't be distorted or disappeared after general image processes like a compression and filtering. In this paper, we present a new digital image watermarking algorithm which combines the discrete wavelet transform (DWT) and the singular value decomposition (SVD). Simulation results show that the newly proposed algorithm is not only robust nevertheless variable attacks like noise, filtering and JPEG compression but also secure in application.

• Proceedings of the Korea Institutes of Information Security and Cryptology Conference
• /
• 2001.11a
• /
• pp.464-468
• /
• 2001

In this paper, we propose a robust watermarking technique that accepts time scaling, pitch shift, add noise and a lot of lossy compression such as MP3, AAC, WMA. The technique is developed based on digital filtering. Being designed according to critical band of HAS (human auditory system), the digital filters nearly affect audio quality. Furthermore, before implementing digital filtering, wavelet transform decomposes the audio signal into several signals that is composed of specific frequencies. Designed digital filters scan the decomposed signal. The designed digital filter, band-stop filter, distorts and eliminates specific frequencies of audio signals. Watermarking detection can be accomplished by FFT (Fast Fourier Transform). Firstly, segments of audio signal are transformed by FFT. Then, the obtained amplitude spectrum by FFT is summed repeatedly. Finally the watermark detector can find filters used to watermark encoding based on eliminating frequencies. The suggested technique can embed 4bits/s in a robust manner.

• Journal of the Korean Society for Industrial and Applied Mathematics
• /
• v.18 no.2
• /
• pp.143-156
• /
• 2014

This paper aims to give a quick view on denoising without comprehensive details. Denoising can be understood as removing unwanted parts in signals and images. Noise incorporates intrinsic random fluctuations in the data. Since noise is ubiquitous, denoising methods and models are diverse. Starting from what noise means, we briefly discuss a denoising model as maximum a posteriori estimation and relate it with a variational form or energy model. After that we present a few major branches in image and signal processing; filtering, shrinkage or thresholding, regularization and data adapted methods, although it may not be a general way of classifying denoising methods.

• The Journal of Korean Institute of Communications and Information Sciences
• /
• v.32 no.5C
• /
• pp.520-529
• /
• 2007

In this paper, we study the characteristics of parameters which are related to performance of MCTF which is a key technique for wavelet-based scalable video coding, and propose an improved MCTF method. The proposed MCTF method adopts the motion estimation of which motion vector field is distributed more uniformly using variable block sizes. By using the proposed method, the decomposition performance of temporal filter is improved, and the energy in high-frequency frames is reduced. It can help the entropy coder to generate lower bitrate. From simulation results, we verify the decomposed energy on high-frequency frame using the proposed method is reduced by 25.86% at the most in terms of variance of the high-frequency frame.

• Proceedings of the IEEK Conference
• /
• 2001.09a
• /
• pp.845-848
• /
• 2001

본 논문에서는 multipath 환경에서 효과적으로 시지연을 추정하기 위한 알고리즘을 제안하였다. 제안된 알고리즘은 전처리 과정으로 웨이블릿 변환을 적용하였으며, 적응 알고리즘으로는 RLS를 계층적인 구조로 나타낸 HRLS(Hierarchical RLS)를 사용하였다. 시지연은 신호 분해과정 이후 각각의 부밴드에서 primary 신호와 reference 신호 사이의 MSE(Mean of Squared Error)를 최소화 시키는 적응 메카니즘을 사용하여 추정하였다. 아울러 모의실험을 통하여 제안된 알고리즘의 성능을 검증하였다.