Jau-Yien Lee

DETERMINING LOCATION AND ORIENTATION OF A LABELLED CYLINDER USING POINT-PAIR ESTIMATION ALGORITHM

In this paper, a new feasible approach is proposed to determine the 3-D location and orientation as well as the radius of a cylinder pasted with a known-height label. The proposed approach uses a r...

Robot location determination in a complex environment by multiple marks

Abstract A new approach of using multiple marks for determining three-dimensional robot location in a complex environment, which may be a composed space or a space with rough terrain, is proposed. One principal idea is to pre-locate a large enough number of marks with different labels at different planar places that at least one mark can be seen by the camera, wherever the robot is in the working space. The other concept is that by integrating the pattern recognition technique with the three-dimensional geometrical transformation method, the system can determine the robot location without the need of fitting any prerequisite alignment. This novel property enables the robot to work in a space with rough terrain. Experimental results from using three marks in a space composed of two warehouses have shown that the location error is less than 3% on average.

The optimal camera geometry and performance analysis of a trinocular vision system

In this paper, the projection location constraint is exploited to reduce ambiguous correspondences in trinocular systems. The inherent matching problems associated with the spurious features produced by overlapping among targets, the periodic structure of objects and the occlusion features were solved. The projection location constraint can be achieved by the deviation-to-distance transformation and the optimal epipolar constraint. The deviation-to-distance transformation transforms the location deviation of the extracted feature on an image plane to a maximum distance between the potential corresponding features and the epipolar line on another image for restricting the searching area of correspondences. The optimal epipolar constraint is obtained by adjusting the camera model before systems start to work. With the deviation-to-distance transformation and the optimal epipolar constraint, the arbitral area for searching the potential corresponding features can be made reasonably small. This not only saves computational time but also results in fewer ambiguous correspondences. In practice, the importance of the paper is that it proposes a very easy adjustment of camera geometry to simplify the most troublesome matching problem encountered in stereo systems. All the discussions and derivations are based on the geometry of the camera model and the mapping between the locations of the 3D features and their projection locations on the image planes. The non-ambiguity probability of the trinocular system is derived theoretically and verified by experimental results. >

Polyhedron reconstruction using three-view analysis

Abstract Based on the three-view analysis, this paper presents a new matching method for polyhedron reconstruction. The reconstruction process is driven by corners and edge line drawing. Under the requirement of the implicit condition of exact image-globe coordinate model, the matching methodology first uses the necessary condition of the corresponding-point pair, called the PCCF (point-correspondence constraint function), to find the second-order candidates of corresponding-corner triplets among the three images from corner-to-corner inspection. It then uses the local connected corner triplets (LCCTs) to select the first-order candidates. Lastly, it examines the real spatial position of each first-order candidate in the global coordinate system to confirm the exact correspondence. The method is simple and reliable and can be applied to any arbitrary configuration of three camera locations. To accomplish partial and complete polyhedron reconstruction, a feature extraction method is presented. The method is fast and most suitable for noisy data. Experimental results are included in the paper to verify our proposed methods. The measurement error is less than 2.2 mm (0.3%) when the distance between the camera and the object is about 700 mm.

Determination of robot location using generalized cylindrical object shapes

Abstract In this paper, a new approach to the determination of robot location using the generalized cylindrical object, such as circular cylinders, elliptic cylinders, triangular cylinders, etc., are proposed. From a monocular image of the object, image processing and numerical analysis techniques are applied to extract the projection characteristics of the generalized cylindrical portion of the object, from which the position and the rotation parameters of a camera-mounted robot can be determined. Owing to the full linearity of the derivation, this approach can achieve high speed requirement. In addition, no prerequisite restriction is imposed on the image-grabbing process. Experimental results show that the location determination time is about 1.2 sec in a 33 MHz IBM compatible PC/AT computer system and the location error is less than 5% in average.

Determining 3-D location parameters of a cylindrical object

For a robot, to fetch the desired object, it is necessary to determine the objects location relative to the robot, and calculate the orientation of the object. Many of these objects are of cylindrical shape, for example, cylindrical containers with harmful liquid on a desk, cylindrical bottles with drink on the shelf, etc. In this paper, a simple geometrical method is proposed to determine the location, the radius and the orientation of a known height cylindrical object in the 3-D space. This object location determination problem is the equivalence of the problem of robot location by the use of the cylindrical object. Here, any object that has a cylindrical portion is defined as the cylindrical object and such objects are frequently seen in real environment. By integrating image processing techniques, 3-D vector analysis and simple algebraic computation, the system can achieve high speed requirement in determining the location and the orientation of the object. The approach has the advantage that it does not require the values of any intrinsic camera parameters. Experimental results for the desired object taken in different positions show that the computing time for location and orientation determination is about 1.2 sec in a 20 MHZ IBM compatible PC/AT computer system and the measurement error is less than 5% on the average.

The Fibonacci search for cornerpoint detection of two-dimensional images

A new algorithm for cornerpoint detection using the Fibonacci search method is derived. The algorithm is an optimization-based unconstrained line search method which can be used to approximate a 2-D non-polygon object shape to any desired accuracy. An adjustment procedure including corner merging and false corner detection is also included. The computation time of our proposed approach is only 10% of that required by other cornerpoint detection methods such as the merging method, the cubic B-spline method and the conventional splitting method.

A fast method for cylindrical object location determination

Abstract In this paper, a simple and fast geometrical single-view method is proposed to determine the location of labelled cylindrical objects in the 3-D space. For a robot, to fetch the desired cylindrical object, two problems have to be solved. They are the identification of the cylindrical object, and the determination of the objects location relative to the robot. In the proposed method, a known-height label is pasted on the cylindrical object, on which an object code is associated for the robot to identify the desired object. For determining the location of the object, three-dimensional vector analysis and simple algebraic computation are applied to achieve the high speed requirements. This approach has the advantage that it needs not to know any intrinsic camera parameter. An analysis of location errors for different positions is included. Simulation results show that the location determination time is about 0.2 sec in a 20 MHZ IBM compatible PC/AT computer system, and the location error is less than 2% in average.

Determining location and orientation of a labelled cylinder using point-pair estimation algorithm

A feasible approach is proposed to determine the 3D location and orientation of a cylinder with a label stuck on it at a known height. The proposed approach uses a rectangle-shaped standard mark composed of two point-pairs for performing monocular image analysis. A monocular image of a labelled cylinder is first taken. Image processing and numerical analysis techniques are then used to select two point-pairs located on the boundary of the visible part of the cylindrical label. According to 3D imaging geometry, the location and orientation of the cylinder relative to the camera are uniquely determined by using 3D vector analysis and simple algebraic computation. Owing to the full linearity of the derivation, this approach can perform at high speed.<<ETX>>

A new approach to robot guidance in unfamiliar environment using an indication post

A novel approach to robot guidance in an unfamiliar environment is presented. In previous guiding methods, a preinstalled map or predefined path is required for a robot navigating in its working space. The present method uses an indication post (IP) to provide the information on the destination, the direction, and the distance from the source point to the destination point, assuming that the mobile robot has no information concerning its workplace. To realize the concept, many IPs depicting simple information were designed. The content of IPs can be described by using a finite-state grammar. Image-processing, statistical, and syntactic pattern recognition approaches are integrated to solve the problems of IP finding, robot location determination, IP identification, and understanding. Experiments simulating practical environments were performed. The results verify that the IP candidates can be automatically found and the depicted information can be extracted and understood correctly, so that the mobile robot is able to plan a global optimal path from an arbitrary start point to an arbitrary end point.<<ETX>>