Jia-Hong Lee

An iterative method of palette-based image steganography

This article presents a novel iterative method of palette-based image steganography that minimizes the RMS error between an original image and its stego-image. The proposed method is based on a palette modification scheme, which can iteratively embed one message bit into each pixel in a palette-based image. In each iteration, both the cost of removing an entry color in a palette and the benefit of generating a new one to replace it are calculated. If the maximal benefit exceeds the minimal cost, a entry color is replaced. Experimental results show that the proposed method remarkably reduces the distortion of the carrier images (stego-images) to other palette-based methods.

Texture classification method using multiple space filling curves

Abstract Texture analysis by using the Peano curve does not provide sufficient discriminatory power to classify natural textures. An improved method by simultaneously using several space filling curves to the texture analysis and classification is proposed.

Texture classification using fuzzy uncertainty texture spectrum

Abstract A new method using fuzzy uncertainty, which measures the uncertainty of the uniform surface in an image, is proposed for texture analysis. A grey-scale image can be transformed into a fuzzy image by the uncertainty definition. The distribution of the membership in a measured fuzzy image, denoted by the fuzzy uncertainty texture spectrum (FUTS), is used as the texture feature for texture analysis. To evaluate the performance of the proposed method. supervised texture classification and rotated texture classification are applied. Experimental results reveal high-accuracy classification rates and show that the proposed method is a good tool for texture analysis.

A novel hybrid learning algorithm for parametric fuzzy CMAC networks and its classification applications

This paper shows fundamentals and applications of the parametric fuzzy cerebellar model articulation controller (P-FCMAC) network. It resembles a neural structure that derived from the Albus CMAC and Takagi-Sugeno-Kang parametric fuzzy inference systems. In this paper, a novel hybrid learning which consists of self-clustering algorithm (SCA) and modified genetic algorithms (MGA) is proposed for solving the classification problems. The SCA scheme is a fast, one-pass algorithm for a dynamic estimation of the number of hypercube cells in an input data space. The clustering technique does not require prior knowledge such as the number of clusters present in a data set. The number of fuzzy hypercube cells and the adjustable parameters in P-FCMAC are designed by the MGA. The MGA uses the sequential-search based efficient generation (SSEG) method to generate an initial population to determine the most efficient mutation points. Illustrative examples were conducted to show the performance and applicability of the proposed model.

Reversible data-hiding method for palette-based images

This study proposes a novel reversible data-hiding method based on an iterative approach for palette-based images. In each iteration, the number of occurrences of palette color in the image is computed in order to identify the most and least frequently occurring colors in a palette for data hiding. Data are hidden by sacrificing the least frequent color and placing the most frequent color in its position in the palette to make a pair . After the most frequent color takes over the palette index of the least frequent color, two palette indices stand for the most frequent color. Therefore, one bit can be hidden in every pixel where the most frequent color occurs in the image. To make the host image reversible, the overhead information caused by sacrificing the least frequent color must be saved for image reconstruction. The performance of the proposed method is demonstrated on test images by showing the capacity and distortion. Several successful similar efforts have also been implemented for comparison. Experimental results indicate that the proposed method has a high embedding capacity and good marked-image quality.

A reversible data-hiding method for palette-based images with capacity optimization

This study presents a reversible data-hiding scheme based on palette image histogram characteristics, in which the least frequent color entries and the most frequent color entries are identified to imbed data. Data are concealed by sacrificing many sets of the least frequent colors, and placing one of the most frequent colors in their positions in the palette to make a ‘cluster’ for each set. Different bits of data can be imbedded using different sized clusters. A capacity optimization scheme is then developed to obtain the maximum capacity by estimating all possible combinations of clusters. For image reconstruction, making the host image reversible requires saving the overhead information caused by sacrificing the least frequent colors. Effectiveness of the proposed method is also demonstrated on test images by showing the capacity and distortion. Importantly, the proposed method has a high imbedding capacity and excellent marked image quality.

A new inverse halftoning method using reversible data hiding for halftone images

A new inverse halftoning algorithm based on reversible data hiding techniques for halftone images is proposed in this paper. The proposed scheme has the advantages of two commonly used methods, the lookup table (LUT) and Gaussian filtering methods. We embed a part of important LUT templates into a halftone image and restore the lossless image after these templates have been extracted. Then a hybrid method is performed to reconstruct a grayscale image from the halftone image. In the image reconstruction process, the halftone image is scanned pixel by pixel. If the scanned pattern surrounding a pixel appeared in the LUT templates, a gray value is directly predicted using the LUT value; otherwise, it is predicted using Gaussian filtering. Experimental results show that the reconstructed grayscale images using the proposed scheme own better quality than both the LUT and Gaussian filtering methods.

Object watermarking scheme based on resynchronization and shape subdivision

In practical use, objects in still images and videos are easy to misappropriate with handy image-processing tools and need to be pro- tected. However, the few proposals for object watermarking suffer from resynchronization problems. In this study, we develop an object-based watermarking method based on efficient segmentation using lines paral- lel and perpendicular to two principal axes in the spatial domain. A patch- like scheme is designed to embed and extract invisible watermarks. In contrast with the previous object-based watermarking schemes, extra information for resynchronization is not required to be stored in our method. Experimental results show the robustness of the proposed method against many kinds of geometrical attacks.

A new method for texture analysis using morphological gradient texture histograms

Abstract A new method using Morphological Gradients, which measures the degree of roughness of image surface, is proposed for texture analysis. In this method, the local texture feature for a given pixel is characterized by its corresponding morphological gradient, and the global texture aspect of an image, denoted the Morphological Gradient Texture Histogram (MGTH), is used as a distinguishing feature for texture analysis and classification. To evaluate the performance of the proposed method, supervised texture classification and rotated texture classification are applied. Experimental results reveal high accuracy classification rates and show that the proposed method is a good tool for texture analysis.

Data Hiding Scheme to Improve the Quality of Reconstructed Images for Inverse Halftoning

The objective of inverse halftone reconstruction is to convert halftone bilevel images into gray-level images with the minimum distortion, but there is no way to construct a perfect gray-scale image from a given halftone image. A reversible data hiding approach is proposed to improve the recovered image quality for halftone images. The proposed approach firstly computes the difference image between the predicted image using Gaussian filtering from a halftone image and the original gray-scale one. The difference image is then compressed and embedded into the halftone image using a reversible data hiding method. In the image reconstruction process, the marked halftone image is scanned to extract the embedded difference image. Finally, a better quality of reconstructed image will be generated by adding the extracted difference image to the predicted image using Gaussian filtering. Experimental results show that the proposed approach can efficiently improve the quality of reconstructed images for inverse halftoning.