Cognex Corporation

Nigel John Foster

William M Silver

E John McGarry

Sanjay J Nichani

Willard P Foster

Adam Wagman

Matthew Leon Hill

Jane Alice Loizeaux

Cyril C Marrion

Yian Leng Chang

Aaron S Wallack

David Y Li

Xiangyun Ye

Lifeng Liu

Guruprasad Shivaram

Matt Hill

This invention provides a system and method for determining the three-dimensional alignment of a modeled object or scene. A 3D (stereo) sensor system views the object to derive a runtime 3D representation of the scene containing the object. Rectified images from each stereo head are preprocessed to enhance their edge features. 3D points are computed for each pair of cameras to derive a 3D point cloud. The amount of 3D data from the point cloud is reduced by extracting higher-level geometric shapes (HLGS), such as line segments. Found HLGS from runtime are corresponded to HLGS on the model to produce candidate 3D poses. A coarse scoring process prunes the number of poses. The remaining candidate poses are then subjected to a further more-refined scoring process. These surviving candidate poses are then verified whereby the closest match is the best refined three-dimensional pose.

System and method for three-dimensional alignment of objects using machine vision

Disclosed is a method for determining the absence or presence of one or more instances of a predetermined pattern in an image, and for determining the location of each found instance within a multidimensional space. A model represents the pattern to be found, the model including a plurality of probes. Each probe represents a relative position at which a test is performed in an image at a given pose, each such test contributing evidence that the pattern exists at the pose. The method further includes a comparison of the model with a run-time image at each of a plurality of poses. A match score is computed at each pose to provide a match score surface. Then, the match score is compared with an accept threshold, and used to provide the location any instances of the pattern in the image.

Method for fast, robust, multi-dimensional pattern recognition

A ball grid array inspection and location method and apparatus includes a raw feature finding processor which uses a feature finding algorithm to find ball features (irrespective of number) and generate a list of raw features as an X and Y location for each feature located. An angle estimation processor receives the list of raw features, a user estimated orientation angle, and ball-to-ball pitch information and determines an estimated grid angle. An on-grid check processor receives the estimated grid angle, the list of raw features and the ball-to-ball pitch information and produces a set or list of on-grid features in the form of a list X and Y translation parameters for each feature/ball found on-grid. A correspondence processor receives the list of on-grid features and model grid information and searches for correspondence between the two dimensional array of found grid features and the two dimensional array of model grid features to yield a direct mapping of respective two dimensional spatial relationships between model features and found features, which is used to determine the spatial relationship between model space and physical space in order to position/align balls of the device in physical space in accordance with model balls in model space. A least square fit is effected to refine position of the device in terms of its translation and rotation.

Method and apparatus for locating ball grid array packages from two-dimensional image data

Method and apparatus for locating ball grid array packages from

The invention provides a two-phase method for identifying sets of collinear points in data points, e.g., derived from an image. During a first phase, the 'angle-finding' phase, the angular orientation of the lines defined by the sets of points is estimated. During a second phase, the 'line-finding' phase, sets of collinear points aligned at that angle are identified. The invention has application, for example, in finding parallel and perpendicular point sets in an image.

Machine vision methods for identifying collinear sets of points from an

A machine vision system for finding an angle object orientation, or other characteristic, of an object in an image. The system determines the boundary pixels, determines the segment orientation angles associated with each successive segment of n boundary pixels, assigns a weight to each angle, generates a histogram of the angles according to their weights, normalizes the histogram to an angle range covering 90.degree., and generates a signal representing the angle corresponding to a selected peak in the histogram. In one embodiment, the system interpolates between neighboring pixels across the boundary to obtain a grey-scale boundary of sub-pixel accuracy.

Machine vision method and apparatus for finding an object orientation

Growing community of inventors

Newton, MA, United States of America

Nigel John Foster