Horace Ho-Shing Ip

Discriminative wavelet shape descriptors for recognition of 2-D patterns

In this paper, we present a set of wavelet moment invariants, together with a discriminative feature selection method, for the classification of seemingly similar objects with subtle di⁄erences. These invariant features are selected automatically based on the discrimination measures defined for the invariant features. Using a minimum-distance classifier, our wavelet moment invariants achieved the highest classification rate for all four di⁄erent sets tested, compared with Zernike’s moment invariants and Li’s moment invariants. For a test set consisting of 26 upper cased English letters, wavelet moment invariants could obtain 100% classification rate when applied to 26]30 randomly generated noisy and scaled letters, whereas Zernike’s moment invariants and Li’s moment invariants obtained only 98.7 and 75.3%, respectively. The theoretical and experimental analyses in this paper prove that the proposed method has the ability to classify many types of image objects, and is particularly suitable for classifying seemingly similar objects with subtle di⁄erences. ( 1999 Pattern Recognition Society. Published by Elsevier Science Ltd. All rights reserved.

Three-dimensional virtual-reality surgical planning and soft-tissue prediction for orthognathic surgery

Complex maxillofacial malformations continue to present challenges in analysis and correction beyond modern technology. The purpose of this paper is to present a virtual reality workbench for surgeons to perform virtual orthognathic surgical planning and soft-tissue prediction in three dimensions. A resulting surgical planning system, i.e., three-dimensional virtual reality surgical planning and soft-tissue prediction for orthognathic surgery, consists of four major stages: computed tomography (CT) data post-processing and reconstruction, three-dimensional (3-D) color facial soft-tissue model generation, virtual surgical planning and simulation, soft-tissue-change preoperative prediction. The surgical planning and simulation are based on a 3D CT reconstructed bone model, whereas the soft-tissue prediction is based on color texture-mapped and individualized facial soft-tissue model. Our approach is able to provide a quantitative osteotomy-simulated bone model and prediction of postoperative appearance with photorealistic quality. The prediction appearance can be visualized from any arbitrary viewing point using a low-cost personal computer-based system. This cost-effective solution can be easily adopted in any hospital for daily use.

Constructing a 3D individualized head model from two orthogonal views

A new scheme for constructing a 3D individualized head model automatically from only a side view and the front view of the face is presented. The approach instantiates a generic 3D head model based on a set of the individuals facial features extracted by a local maximum-curvature tracking (LMCT) algorithm that we have developed. A distortion vector field that deforms the generic model to that of the individual is computed by correspondence matching and interpolation. The input of the two facial images are blended and texture-mapped onto the 3D head model. Arbitrary views of a person can be generated from two orthogonal images and can be implemented efficiently on a low-cost, PC-based platform.

Symmetry detection by generalized complex (GC) moments: a close-form solution

This paper presents a unified method for detecting both reflection-symmetry and rotation-symmetry of 2D images based on an identical set of features, i.e., the first three nonzero generalized complex (GC) moments. This method is theoretically guaranteed to detect all the axes of symmetries of every 2D image, if more nonzero GC moments are used in the feature set. Furthermore, we establish the relationship between reflectional symmetry and rotational symmetry in an image, which can be used to check the correctness of symmetry detection. This method has been demonstrated experimentally using more than 200 images.

Virtual brush: a model-based synthesis of Chinese calligraphy

Abstract This paper describes the Virtual Brush, a model-based approach to synthesizing realistically Chinese calligraphic writings. This approach simulates the physical process of brush stroke creation using a parameterized model which captures (a) the writing brush 3D geometric parameters, (b) the brush hair properties and (c) the variations of ink deposition along a stroke trajectory. An analysis of some well-known Chinese brush writing styles is given, with a view to establishing the relationship of these writing styles and the modeling process of simulating these different writing styles. We present here our model formulation and the implementation of a software which is capable of simulating some typical calligraphic writing styles on a PC platform. This parameterized model allows a compact representation for brush-written characters which can be used to synthesize characters at different levels of detail, with various brush effects. The result is applicable for very high-quality publishing purpose. Additionally, by appropriately transforming and scaling the stroke trajectories which define a virtual brush character, this technique can be used as a novel approach for scalable font design and generation, in comparison with traditional vector fonts or imaging techniques. More interestingly, with suitable interfacing techniques Virtual Brush would allow users to “practice” calligraphy electronically.

An affine-invariant active contour model (AI-snake) for model-based segmentation

In this paper, we show that existing shaped-based active contour models are not affine-invariant and we addressed the problem by presenting an affine-invariant snake model (AI-snake) such that its energy function are defined in terms local and global affine-invariant features. The main characteristic of the AI-snake is that, during the process of object extraction, the pose of the model contour is dynamically adjusted such that it is in alignment with the current snake contour by solving the snake-prototype correspondence problem and determining the required affine transformation. In addition, we formulate the correspondence matching between the snake and the object prototype as an error minimization process between two feature vectors which capture both local and global deformation information. We show that the technique is robust against object deformations and complex scenes.

Affine-invariant image retrieval by correspondence matching of shapes

In this paper, we present an efficient multi-scale similarity matching method for shape-based image indexing and retrieval. This method is affine-invariant and stable against noise and shape deformations. Shapes which have undergone mirror reflection can also be retrieved in a unified manner. In this approach, similarity matching is cast as correspondence matching of two shapes which is then solved by minimizing the matching errors between two feature vectors. Since our feature vectors simultaneously capture both local and global affine-invariant features of shapes, this formulation makes our solution to the correspondence problem very robust. To render the technique suitable for interactive image retrieval, a fast error minimization algorithm for computing correspondence matching is further proposed. Theoretical analysis and experimental results show that multi-scale similarity matching allows dissimilar shapes to be filtered out very quickly and the resulting method meets the performance and flexibility needed for content-based image indexing and retrieval.

A robust watermarking scheme for 3D triangular mesh models

Copyright protection of digital media has become an important issue in the creation and distribution of digital content. As a solution to this problem, digital watermarking techniques have been developed for embedding specific information identifying the owner in the host data imperceptibly. Most watermarking methods developed to date mainly focused on digital media such as images, video, audio, and text. Relatively few watermarking methods have been presented for 3D graphical models. In this paper we propose a robust 3D graphical model watermarking scheme for triangle meshes. Our approach embeds watermark information by perturbing the distance between the vertices of the model to the center of the model. More importantly, to make our watermarking scheme robust against various forms of attack while preserving the visual quality of the models our approach distributes information corresponds to a bit of the watermark over the entire model, and the strength of the embedded watermark signal is adaptive with respect to the local geometry of the model. We also introduce a weighting scheme in the watermark extraction process that makes watermark detection more robust against attacks. Experiments show that this watermarking scheme is able to withstand common attacks on 3D models such as mesh simplification, addition of noise, model cropping as well as a combination of these attacks.

A review of intelligent content-based indexing and browsing of medical images

Physicians are beginning to be able to gain access, through the Internet, to the world’s collections of multimedia medical information such as MRI (magnetic resonance imaging) and CT (computer tomography) image archives, videos of surgical operations and medical lectures, textual patient records and media annotations. New techniques and tools are needed to represent, index, store and retrieve digital content efficiently across large collections. In this review, we trace the development of visual information systems for healthcare and medicine from Picture Archiving and Communications Systems (PACS) to the recent advances in content-based image retrieval, whereby images are retrieved based on their visual content similarity - that is, colour, texture, and shape. Medical images, unlike consumer-oriented images, pose additional challenges to content-based image retrieval, in that visual features of normal and pathological images are typically separated by only subtle differences in visual appearance. Intelligent image retrieval and browsing therefore requires a combination of prior knowledge of the medical domain, image content and image annotation analysis. To this end, we also overview the I-Browse project, conducted jointly by the Clinical School of the University of Cambridge and the City University of Hong Kong, which aims to develop techniques which enable a physician to search over image archives through a combination of semantic and iconic contents.

Constrained spectral clustering via exhaustive and efficient constraint propagation

This paper presents an exhaustive and efficient constraint propagation approach to exploiting pairwise constraints for spectral clustering. Since traditional label propagation techniques cannot be readily generalized to propagate pairwise constraints, we tackle the constraint propagation problem inversely by decomposing it to a set of independent label propagation subproblems which are further solved in quadratic time using semi-supervised learning based on k-nearest neighbors graphs. Since this time complexity is proportional to the number of all possible pairwise constraints, our approach gives a computationally efficient solution for exhaustively propagating pairwise constraint throughout the entire dataset. The resulting exhaustive set of propagated pairwise constraints are then used to adjust the weight (or similarity) matrix for spectral clustering. It is worth noting that this paper first clearly shows how pairwise constraints are propagated independently and then accumulated into a conciliatory closed-form solution. Experimental results on real-life datasets demonstrate that our approach to constrained spectral clustering outperforms the state-of-the-art techniques.