Laser dark field inspection tools for LSI wafer patterns are very fast and often used in LSI fabrications. However with non-repetitive LSI patterns, the sensitivity of the conventional die-to-die inspection method is not sufficient due to the increase in noise that occurs in connection with highly coherent laser illumination and the variation of the thin film thickness of the LSI patterns. In this paper two methods are combined for highly sensitive image comparison : self-reference comparison and die-to-golden die comparison. The self-reference comparison searches for parts of similar patterns in each die and the die-to-golden die comparison generates images with similar patterns on the basis of adjacent dies. The combined methods adaptively select the parts to be referenced pixel by pixel. Experimental results showed they could successfully detect a low S/N defect signal of 0.12 in the non-repetitive patterns.
This paper describes a method to solve the binocular stereo correspondance problem on an epipolar plane. In porposed method, a parametric curve is on the epipolar plane which is calcurated using world coordinates of lens centers and a image point. And a coordinate frame is set up on the epipoar plane(shn), and a transformation matrix between the world coordinate system and the coordinate frame (shn) is calcurated. Coordinates of the parametric curve is transformed from the coordinate frame (shn) to the world coordinate system. The parametric curve is projected to two cameras, and these projected lines have correspondance area on the images. Pixel values in area around these correspondance line on each images are compared to calcurate errors. And an error-function is defined the errors. Parameters of the curve are optimzed to let the error function be smaller. The laid parameteric curve is a polynominal and the polytope method is used as optimizing method. Pixel values of epipolar line on each two images are fitted each other on the images. A board shape object and a column shape object are used as experimentation. And the shape of the parametric curve is shown as a result.
In recent years, the CMM (Coordinate Measuring Machine) quickly has been spreading in the manufacturing industry. With this phenomenon, the environmental condition and installation site of CMMs has been changed from the temperature control room to the workshop environment and the production line without air conditioning. However, even in severe environment the demand to high accuracy measurements by CMM has been enhancing still more. Therefore, the methods to evaluate and realize the high accuracy measurements by CMM in workshop environment have been investigated. In this paper, we evaluated the influence of measurement error of indication within form, size and location errors by articulating probing system. Furthermore, we experimentally tested the influence of measurement error of indication within form, size and location errors, and proposed the method to evaluate location errors with the value of the squareness error of the position of the reference sphere and the position of measurement.
Sensing in aquatic environments is important to maintain underwater structures and research aquatic life. This paper proposes a 3D measurement method of objects in water using a fish-eye stereo camera. Aquatic sensing meets the difficulty of 3D measurement because of light refraction due to the difference of refractive indices of air and water. This problem is enhanced by distortion of fish-eye images. The proposed method removes distortion of fish-eye images by converting them to perspective projection images. The corrected images are searched for corresponding points and measurement points are located by triangulation. The effect of light refraction is considered by ray tracing to measure accurately in aquatic environments. Experimental results show the effectiveness of the proposed method.
This paper proposes a new navigation method that makes a mobile robot move without map information. Past navigation methods need map information and the robot's self-location. If the robot has no information of map or mistakes localization, the navigation method cannot be used. To overcome this problem, we introduce a new concept “leading vector” that indicates about the direction of the destination (local route information). The robot can move smoothly by using the leading vector if the passage structure is not complicated. However, the robot may be in danger of collision in compli-cated environments only using the leading vector. Therefore, the proposed method uses pedestrians' flow information. The environmental information is enclosed by the pedestrians' flow information at second hand. The robot can reach its destination by utilizing the leading vector using the local route information and the pedestrians' flow information. In this paper, simulation experiment was conducted, and its results show the effectiveness of the proposed method.
For practical application, 3D-MEMS optical switch modules should be a mechanical-vibration-proof and have good temperature characteristics and good electromagnetic compatibility (EMC). To satisfy these fundamental environmental characteristics of optical switch modules, we examine a design technique of vibration-proof mechanism, a structure of fixing mechanism for optical I/O device and a module structure considered for EMC performances. We evaluated the fundamental environmental characteristics of a prototype optical switch module. In results, we confirmed that the switching module satisfies the Telcordia GR-63 office vibration test due to application of vibration-proof mechanism. The structure of fixing mechanism for optical I/O devices enabled the reduction of error in optical axis alignment. We also show that the CISPR 22 and CISPR24 are satisfied for EMC performance in this prototype. The evaluation results confirm that the 3D MEMS optical switch module is suitable for practical use.
Profile errors of grooves milled with a micro square end mill are discussed in this paper. Very low stiffness of micro end mills causes inevitably those errors to depress the productivity. Upright errors on both side walls of a groove are caused by transient chip formation on twisted cutting edges, that is, cutting point moves from the groove bottom to the shoulder at an engaging cross section of a groove, where uncut chip gets gradually larger in width. In the other hand, uncut chip width gets gradually smaller at disengaging cross section of a groove. Thus changing chip volume at both side walls forces the change of elastic deflection by the difference of normal cutting force component on both cutting edges. It results in upright errors on both side walls. Additionally, the end face of an end mill is inclined by normal cutting force, and it makes the concave bottom surface with an ellipse sloping down to a down-milling wall, because the largest cutting force toward the tool axis acts on a cutting edge at a rotating angle over 90deg. from engaging point.
High Young's modulus steel containing TiB2 particles is known as one of the difficult-to-cut materials. For this reason, a tool life becomes very short in machining of this material. The purpose of this study is to investigate extension of the tool life in the turning of high Young's modulus steel. The optimum cutting speed and cutting edge shape were examined from the viewpoint of tool life and surface roughness. The results indicated a dramatic decrease in the tool life of K10-C tool due to scratching of the flank wear region by hard TiB2 particles. On the other hand, the cutting edge of a binder less cBN tool with a negative land angle was protected by the built-up edge, leading to a decrease in the cutting force and an extension of the tool life. For this tool, it was clarified that from the viewpoint of wear resistance and the machined surface, the optimum conditions were a cutting speed of 1.0 m/s and a negative land angle of 15 degrees. It was shown that the presence of a secondary chip allows these conditions to be achieved, and that a built-up edge leads to stability.