In this Book we argue that the fruitful interaction of computer vision and belief calculus is capable of stimulating significant advances in both fields. From a methodological point of view, novel theoretical results concerning the geometric and algebraic properties of belief functions as mathematical objects are illustrated and discussed in Part II, with a focus on both a perspective 'geometric approach' to uncertainty and an algebraic solution to the issue of conflicting evidence. In Part III we show how these theoretical developments arise from important computer vision problems (such as articulated object tracking, data association and object pose estimation) to which, in turn, the evidential formalism is able to provide interesting new solutions. Finally, some initial steps towards a generalization of the notion of total probability to belief functions are taken, in the perspective of endowing the theory of evidence with a complete battery of estimation and inference tools to the benefit of all scientists and practitioners.

Visions of a Generalized Probability Theory

A health monitoring method using a tongue image including constructing a database in which a tongue image obtained from a person, a template image set for the person, and a result of determining a health condition of the person with respect to at least one characteristic factor are linked together; isolating a region of interest within the tongue image, which is acquired from a person whose health condition is to be determined, using template matching; detecting at least one characteristic factor from the isolated region of interest; and determining a health condition of the person based on a change in an appearance of the tongue, which is detected by a comparison between the detected characteristic factor and a characteristic factor searched from the database with respect to the region of interest in different health conditions.

Method of digital image analysis for isolating a region of interest within a tongue image and health monitoring method and apparatus using the tongue image

This paper presents an application of Kohonen's self-organizing feature maps (SOM) for solving the problem of color constancy. The main problem is to evaluate the transformation between collections of colorpoints forming differently shaped clouds in color space under changing illumination. The main idea is to embed appropriate 3D coordinate systems into these clouds by self-organization, and so to be able to find corresponding color points within different clouds. The difference of the locations of corresponding neurons in two SOMs is then an approximation for the particular color shift belonging to the difference between a reference illumination and a given illumination. The observed shifts provide a table of vectors in the color space, which in correction steps can be applied to color images taken under a given illuminant.

Color-Calibration of Robot Vision System Using Self-Organizing Feature Maps

A method of extracting a region of interest from a tongue image and a health monitoring method and apparatus using the tongue image are provided. The health monitoring method includes constructing a database in which a tongue image obtained from a person, a template image set for the person, and a result of determining a health state of the person with respect to at least one characteristic factor are linked to one another; extracting a region of interest from a tongue image, which is acquired from a person whose health state is to be determined, using template matching; detecting at least one characteristic factor from the extracted region of interest; and determining the person's health state based on a change in a tongue state, which is detected from a comparison between the detected characteristic factor and a characteristic factor searched from the database with respect to the region of interest in different health states.

Method of extracting region of interest from tongue image and health monitoring method and apparatus using the tongue image

In this paper, we present a new method for reconstruction of refractive index field of a transparent medium from multi-frame intcrfcrograms obtained by several interferometers in different viewing directions. It is based on double cubic many-knot interpolating splines and A RT techniques. It can offer significant improvement in the calculational accuracy and combine various constrains. The details of the method are discussed in the paper. Experimental results show that the method is effective.

Reconstruction Of 3-D Refractive Index Fields From Multi-Frame Interfetometric Data

This presentation discusses the problem of segmentation of nuclei in cytological color images in different color spaces, namely RGB and HSI color spaces, for the detection of lung cancer cells. For the segmentation in each color space, the background and foreground of the images are first defined, and the chromatic mean values of the background and foreground are then extracted. In the learning phase, based on the chromatic mean values of the background and foreground of training samples, an adaptive threshold function is constructed for each color space using the B-Spline technique. The nuclei are then segmented by thresholding using the adaptive threshold function obtained in the learning phase. Comparisons between the segmentation in RGB color space and in HSI color space are carried out.

Efficient segmentation of nuclei in different color spaces

Data fusion has already been widely used in various applications in which multiple sources of information are presented. One of the most widely used applications of data fusion is in the field of object recognition and classification, since it can efficiently improve the accuracy and the ability of fault tolerance. This presentation describes a cytological color image processing system, using data fusion method, developed for the detection of early stage lung cancers used in the health inspection. As most of the existing microscopic diagnostic systems use morphological, textual and gray or color features respectively, which results in the instability of the diagnosis, this system makes use of all these features by fusing multiple classification results obtained using morphological, textual and chromatic features respectively. Data fusion is achieved using Dempster-Shafer's Evidential Reasoning (DSER). In the current system, all the nuclei are first segmented by thresholding in a special color space which is a non-linear transformation of the (R,G,B) color space. Then, using morphological, textual and chromatic features respectively, the segmented cells are classified as normal or abnormal cells. Using DSER, the classification results obtained above are fused into a final result. Finally, a decision strategy based on the fused data is presented to get the final classification results. And experiment results are given to show the feasibility of the data fusion approach proposed here.© (1993) COPYRIGHT SPIE--The International Society for Optical Engineering. Downloading of the abstract is permitted for personal use only.

Data fusion for detection of early-stage lung cancer cells using evidential reasoning

In practical applications, it is possible to acquire multidirectional interferometric data of different instants. In such cases, refractive-index fileds can be reconstructed based on the multidirectional interferometric data of two consecutive times. The advantages of using the interferometric data of different instants include: more a priori information and knowledge about refractive-index fields can be incorporated in reconstruction and the problem of sovling velociety, temperature and pressure fields will be simplified. In this paper. An iterative direct 3D reconstruction algorithm from the interferometric data of two consecutive times is presented. The problem of calculating physical fields of fluid is discussed. The experimental results are also provided.

Reconstruction of physical fields of fluid based on interferometric image sequences

mathematical viewpoint, direct 3—D image reconstruction techniques are superior to the 3—D reconstructiontechniques based on reconstructing cross— sectional slices of an object body in the cases : (1) all the projectiondata must be considered together (e.g., cone—beam data); and (2) 3—D reconstruction from incomplete data in

Direct 3D reconstruction based on many-knot interpolating spline technique

A direct 3-D method is proposed in this paper for reconstructions of strongly refractive index fields. The basic idea is to reconstruct a refractive index field and to identify the ray path simultaneously. The 3-D ray tracing method and the curved ray algebraic inversion method based on many-knot interpolating spline technique and the solving method for linear equations by bidiagonalizing matrices are presented to realize the idea. The simulation experimental results are also provided.© (1992) COPYRIGHT SPIE--The International Society for Optical Engineering. Downloading of the abstract is permitted for personal use only.

Jian Lu

Papers

Efficient segmentation of nuclei in different color spaces

Data fusion for detection of early-stage lung cancer cells using evidential reasoning

Reconstruction of physical fields of fluid based on interferometric image sequences

Direct 3D reconstruction based on many-knot interpolating spline technique

Tomographic reconstruction of strongly refracting fields and identification of light rays simultaneously