Jean Devars

THE PERSPECTIVE-N-POINT PROBLEM FOR CATADIOPTRIC SENSORS: AN ANALYTICAL APPROACH

The perspective-N-point problem is a well known issue in computer vision. It consists in the determination of the distance between the camera and a set of points well known in an object coordinate space. This problem has been ex- tensively treated in the literature and is still opened. Many solutions already exist. All these approaches consider only common planar camera. We propose, with a new formulation, to extend this problem to non linear imaging sensors: catadioptric panoramic sensors. The proposed approach permits to get a strictly analytical solution to the perspective-N-point problem usable with this kind of sensors.

AN ANALYTICAL SOLUTION TO THE PERSPECTIVE-N-POINT PROBLEM FOR COMMON PLANAR CAMERA AND FOR CATADIOPTRIC SENSOR

The Perspective-N-Point problem (PNP) is a notable problem in computer vision. It consists of given N points known in an object coordinate space and their projection onto the image, estimating the distance between the video camera and the set of points. By the use of an unusual formulation, we propose a method to get a strictly analytical solution based on the resolution of linear systems. This solution can be computed instantly and is well adapted to real time computer vision applications. Our approach is general enough to work with a nonlinear sensor like a catadioptric panoramic sensor. To improve the localization accuracy, we also provide a technique to correct geometrical distortion. This algorithm also corrects little errors on intrinsic and extrinsic parameters. Well implemented, this correction can be performed in real time.

Object image retrieval with image compactness vectors

We present in this paper a new global measure to characterise an image: the compactness vector. This measure considers both object shape and grey level distribution function and does not require any preliminary segmentation. It is invariant to rotation, translation, scale and luminance and is then a powerful tool for image retrieval from a query image. We present here some object retrieval examples from large database images.

A local color descriptor for efficient scene-object recognition

This paper presents an effective use of local descriptors for object or scene recognition and indexing. This approach is in keeping with model-based recognition systems and consists of an extension of a standard point-to-point matching between two images. Aiming at this, we address the use of Full-Zernike moments as a reliable local characterization of the image signal. A fundamental characteristic of the used descriptors is then their ability to absorb a given set of potential image modifications. Their design calls principally for the theory of invariants. A built-in invariance to similarities allows one to manage narrow bounded perspective transformations. Moreover we provide a study of the substantial and costless contribution of the use of color information. In order to achieve photometric invariance, different types of normalization are evaluated through a model-based object recognition task.

An invariant local vector for content-based image retrieval

In this paper, we present the use of Full-Zernike moments as a local characterization of the image signal. Their computation allows us to construct a locally invariant vector, of which the projection in an index table provides a vote for some model-image. This approach is based on the quasi-invariant theory applied to perspective transformation. Then it requires a characterization being invariant to translation, rotation and change of scale in the image; in other respect, an appropriate normalization of the signal delivers an invariance to illuminance conditions.

A Simple and Accurate Camera Calibration for the F180 RoboCup League

Camera calibration is a very important issue in computer vision each time extracting metrics from images is needed. The F180 camera league offers an interesting problem to solve. Camera calibration is needed to locate robots on the field with a very high precision. This paper presents a method specially created to easely calibrate a camera for the F180 league. The method is easy to use and implement, even for people not familiar with computer vision. It gives very acurate and efficient results.

Localization of Robots in F180 League Using Projective Geometry

The F180 RoboCup league relies on a single camera mounted on top of the field. It is of great importance to use an adapted calibration method to locate robots. Most of the methods used are developped for specific application where 3D is required. This paper presents a new calibration method specially developped for the F180 league geometry, allowing the determination of the camera pose parameters and the correction of the parallax in the image due to different heights of observed robots. This method needs one calibration plane that also could be used for correcting optical distortions introduced by the lens.