Catálogo de publicaciones - libros

The analysis of skin color with makeup is needed to examine the effect of makeup to human skin. Foundation is cosmetics to cover undesirable color on skin and gives basic color to skin. It is important evaluate a change of skin color by the foundation. The present paper modeled the optics of skin with foundation makeup by two layers. Then, the surface spectrum of skin with foundation makeup is estimated by the Kubelka-Munk theory for radiation transfer in the turbid medium. The proposed algorithm can predict the spectral shape of skin surface with different types of foundation by appropriately determining model parameters. As an application of the skin reflectance estimation, we have developed a color simulator for human skin with foundation makeup. Experimental results showed a good accuracy of our estimation results.

The use of spectrophotometers for color measurements on printed substrates is widely spread among paper producers as well as within the printing industry. Spectrophotometer measurements are precise, but time-consuming procedures and faster methods are desirable. The rapid development of digital cameras has opened the possibility to use consumer digital cameras as substitutes for spectrophotometers for certain applications such as production control. Two methods for estimating the reflectance properties of objects from camera RGB measurements using linear estimation techniques combined with linear and non-linear constraints are presented. In the experiments, we have investigated if these techniques can be used to measure the reflectance properties of flat objects such as printed pages of paper. Reflectances were converted to CIELAB color values, and the minimization of color errors were evaluated with CIE color difference formulas. Our experiments show that a consumer digital camera can be used as a fast and inexpensive alternative to spectrophotometers for color measurements on printed substrates.

The local binary pattern approach has evolved to represent a significant breakthrough in texture analysis, outperforming earlier methods in many applications. Perhaps the most important property of the LBP operator in real-world applications is its tolerance against illumination changes. Another equally important is its computational simplicity, which makes it possible to analyze images in challenging real-time settings. Recently, we have begun to study image analysis tasks which have not been generally considered texture analysis problems. Our excellent results suggest that that texture and the ideas behind the LBP methodology could have a much wider role in image analysis and computer vision than was thought before.

Several novel methods based on locally extracted object features and spatial constellation models have recently been introduced for invariant object detection and recognition. The accuracy and reliability of the methods depend on the success of both tasks: evidence extraction and spatial constellation model search. In this study an accurate and efficient method for evidence extraction is introduced. The proposed method is based on simple Gabor features and their statistical ranking.

We propose a dynamic visual learning method that aims to identify people by using sparsely distributed multiple surveillance cameras. In the proposed method, virtual viewpoint images are synthesized by interpolating the sparsely distributed images with a simple 3D shape model of the human head, so that virtual densely distributed multiple images can be obtained. The multiple images generate an initial eigenspace in the initial learning step. In the following additional learning step, other distributed cameras capture additional images that update the eigenspace to improve the recognition performance. The discernment capability for personal identification of the proposed method is demonstrated experimentally.

The aim of the study was to develop methods for moving vehicle tracking in aerial image sequences taken over urban areas. The first image of the sequence was manually registered to a map. Corner points were extracted semi-automatically, then tracked along the sequence, to enable video stabilisation by homography estimation. Moving objects were detected by means of adaptive background subtraction. The vehicles were identified among many stabilisation artifacts and tracked, with a simple tracker based on spatiotemporal connected components analysis. While the techniques used were basic, the results turned out to be encouraging, and several improvements are under scrutiny.

In this paper, we present a novel model for representing facial feature point tracks during an facial expression. The model is composed of a static shape part and a time-dependent expression part. We learn the model by tracking the points of interest in video recordings of trained actors making different facial expressions. Our results indicate that the proposed sum of two linear models – a person-dependent shape model and a person-independent expression model – approximates the true feature point motion well.

Our research projects aim to connect various VR/MR platforms and to realize the seamless collaborative works with the intuitive operations whenever and wherever users like. In order to realize an ideal collaborative workspace, an effective handling scheme for both interactive virtual objects and system operation of VR/MR platform is needed. In this paper, we describe the three components, which are the direct manipulation interfaces for potable MR space and for fixed shared VR/MR space, and Interactive 3D Marker for displaying and manipulating virtual objects reflected lighting and geometry conditions of the real world. These components can give users seamless manipulation workspace by the natural operations such as they behave in their daily life. We also introduce Interactive Interior Design System using these components. This system realizes to coordinate the suitable interiors in the real space, by the free manipulation of virtual interior objects reflected the geometry and the lighting condition of the real space.

Estimating camera position and posture can be applied to the fields of augmented reality and robot navigation. In these fields, to obtain absolute position and posture of the camera, sensor-based methods using GPS and magnetic sensors and vision-based methods using input images from the camera have been investigated. However, sensor-based methods are difficult to synchronize the camera and sensors accurately, and usable environments are limited according to selection of sensors. On the other hand, vision-based methods need to allocate many artificial markers otherwise an estimation error will accumulate. Thus, it is difficult to use such methods in large and natural environments. This paper proposes a vision-based camera position and posture estimation method for large environments, which does not require sensors and artificial markers by detecting natural feature points from image sequences taken beforehand and using them as landmarks.

In this talk I will present geometrical computer vision from its early beginnings in projective geometry and photogrammetry to new research on methods in algebraic geometry. In the talk I will give examples of theory for minimal structure and motion problems (central and non-central cameras, 1D and 2D retina), critical configurations and geometry in general as well as practical results of using such theory in 3D reconstruction, navigation, modelling and image interpretation.