A Graph-Based Approach for Making Consensus-Based Decisions in Image Search and Person Re-Identification.

Image matching and retrieval is the underlying problem in various directions of computer vision research, such as image search, biometrics, and person re-identification. The problem involves searching for the closest match to a query image in a database of images. This work presents a method for generating a consensus amongst multiple algorithms for image matching and retrieval. The proposed algorithm, Shortest Hamiltonian Path Estimation (SHaPE), maps the process of ranking candidates based on a set of scores to a graph-theoretic problem. This mapping is extended to incorporate results from multiple sets of scores obtained from different matching algorithms. The problem of consensus-based decision-making is solved by searching for a suitable path in the graph under specified constraints using a two-step process. First, a greedy algorithm is employed to generate an approximate solution. In the second step, the graph is extended and the problem is solved by applying Ant Colony Optimization. Experiments are performed for image search and person re-identification to illustrate the efficiency of SHaPE in image matching and retrieval. Although SHaPE is presented in the context of image retrieval, it can be applied, in general, to any problem involving the ranking of candidates based on multiple sets of scores.

Inferior Breast-Chest Contour Detection in 3-D Images of the Female Torso.

Stereophotogrammetry is finding increased use in clinical breast surgery, both for breast reconstruction after oncological procedures and cosmetic augmentation and reduction. The ability to visualize and quantify morphological features of the breast facilitates pre-operative planning and post-operative outcome assessment. The contour outlining the lower half of the breast is important for the quantitative assessment of breast aesthetics. Based on this inferior breast contour, relevant morphological measures, such as breast symmetry, volume, and ptosis, can be determined. In this paper, we present an approach for automatically detecting the inferior contour of the breast in 3D images. Our approach employs surface curvature analysis and is able to detect the breast contour with high accuracy, achieving an average error of 1.64 mm and a dice coefficient in the range of 0.72-0.87 when compared with the manually annotated contour (ground truth). In addition, the detected contour is used to facilitate the detection of the lowest visible point on the breast, which is an important landmark for breast morphometric analysis.

3D Face Discriminant Analysis Using Gauss-Markov Posterior Marginals.

We present a Markov Random Field model for the analysis of lattices (e.g., images or 3D meshes) in terms of the discriminative information of their vertices. The proposed method provides a measure field that estimates the probability of each vertex being "discriminative" or "nondiscriminative" for a given classification task. To illustrate the applicability and generality of our framework, we use the estimated probabilities as feature scoring to define compact signatures for three different classification tasks: 1) 3D Face Recognition, 2) 3D Facial Expression Recognition, and 3) Ethnicity-based Subject Retrieval, obtaining very competitive results. The main contribution of this work lies in the development of a novel framework for feature selection in scenaria in which the most discriminative information is smoothly distributed along a lattice.

Automated analysis of zebrafish images for screening toxicants.

An important factor facilitating the application of zebrafish in biomedical research is high throughput screening of vertebrate animal models. For example, being able to model the growth of blood vessel in the vasculature system is interesting for understanding both the circulatory system in humans, and for facilitating large scale screening of the influence of various chemicals on vascular development. Compared to other models, the zebrafish embryo is an attractive alternative for environmental risk assessment of chemicals since it offers the possibility to perform high-throughput analyses in vivo. However the lack of an automated image analysis framework restricts high throughput screening. In this paper, we provide a method for quantitative measurements of zebrafish blood vessel morphology since it is difficult to assess changes in vessel structure by visual inspection. The method presented is generalized, i.e. it is not restricted to any specific chemically treated zebrafish, and can be used with wide variety of chemicals.

Embedding topic discovery in conditional random fields model for segmenting nuclei using multispectral data.

Segmentation of cells/nuclei is a challenging problem in 2-D histological and cytological images. Although a large number of algorithms have been proposed, newer efforts continue to be devoted to investigate robust models that could have high level of adaptability with regard to considerable amount of image variability. In this paper, we propose a multiclassification conditional random fields (CRFs) model using a combination of low-level cues (bottom-up) and high-level contextual information (top-down) for separating nuclei from the background. In our approach, the contextual information is extracted by an unsupervised topic discovery process, which efficiently helps to suppress segmentation errors caused by intensity inhomogeneity and variable chromatin texture. In addition, we propose a multilayer CRF, an extension of the traditional single-layer CRF, to handle high-dimensional dataset obtained through spectral microscopy, which provides combined benefits of spectroscopy and imaging microscopy, resulting in the ability to acquire spectral images of microscopic specimen. The approach is evaluated with color images, as well as spectral images. The overall accuracy of the proposed segmentation algorithm reaches 95% when applying multilayer CRF model to the spectral microscopy dataset. Experiments also show that our method outperforms seeded watershed, a widely used algorithm for cell segmentation.

3D imaging for quantitative assessment of toxicity on vascular development in zebrafish.

In this study, we describe the utility of the zebrafish model of in-vivo blood vessel formation as a tool for chemical risk assessment. Time-lapse confocal imaging of embryonic vasculature in the zebrafish is used in conjunction with digital image analysis to monitor and quantify the effect of toxins on vascular development. Non-rigid registration is used to capture changes in vascular morphology over time. Vascular formation in healthy normal and arsenic treated embryos was evaluated for differences in vascular structure using the algorithms developed. Although, the temporal progression of vascular development was similar, significant differences were observed in vessel structure between the toxin treated and healthy fish. This study revealed, for the first time, that vital vascular structures in fish maybe affected by exposure to arsenic. This technique allowed visualization of vascular abnormalities in embryos showing no external signs of malformations.

Automated spatial alignment of 3D torso images.

This paper describes an algorithm for automated spatial alignment of three-dimensional (3D) surface images in order to achieve a pre-defined orientation. Surface images of the torso are acquired from breast cancer patients undergoing reconstructive surgery to facilitate objective evaluation of breast morphology pre-operatively (for treatment planning) and/or post-operatively (for outcome assessment). Based on the viewing angle of the multiple cameras used for stereophotography, the orientation of the acquired torso in the images may vary from the normal upright position. Consequently, when translating this data into a standard 3D framework for visualization and analysis, the co-ordinate geometry differs from the upright position making robust and standardized comparison of images impractical. Moreover, manual manipulation and navigation of images to the desired upright position is subject to user bias. Automating the process of alignment and orientation removes operator bias and permits robust and repeatable adjustment of surface images to a pre-defined or desired spatial geometry.

Multispectral microscopy and cell segmentation for analysis of thyroid fine needle aspiration cytology smears.

This paper discusses the needs for automated tools to aid in the diagnosis of thyroid nodules based on analysis of fine needle aspiration cytology smears. While conventional practices rely on the analysis of grey scale or RGB color images, we present a multispectral microscopy system that uses thirty-one spectral bands for analysis. Discussed are methods and results for system calibration and cell delineation.