Abstract: A system and method for training multiple pattern recognition and registration models commences with a first pattern model. The model is trained from multiple images. Composite models can be used to improve robustness or model small differences in appearance of a target region. Composite models combine data from noisy training images showing instances of underlying patterns to build a single model. A pattern recognition and registration model is generated that spans the entire range of appearances of the target pattern in the set of training images. The set of pattern models can be implemented as either separate instances of pattern finding models or as a pattern multi-model. The underlying models can be standard pattern finding models or pattern finding composite models, or a combination of both.

Abstract: A system and method for training multiple pattern recognition and registration models commences with a first pattern model. The model is trained from multiple images. Composite models can be used to improve robustness or model small differences in appearance of a target region. Composite models combine data from noisy training images showing instances of underlying patterns to build a single model. A pattern recognition and registration model is generated that spans the entire range of appearances of the target pattern in the set of training images. The set of pattern models can be implemented as either separate instances of pattern finding models or as a pattern multi-model. The underlying models can be standard pattern finding models or pattern finding composite models, or a combination of both.

Abstract: This invention provides a system and method for aligning first three-dimensional (3D) point cloud image representing a model with a second 3D point cloud image representing a target, using a vision system processor. A passing overall score is established for possible alignments of the first 3D point cloud image with the second 3D point cloud image. A coverage score for at least one alignment of the first 3D point cloud image with the second 3D point cloud image is estimated so that the coverage score describes an amount of desired features in the first 3D point cloud image present in the second 3D point cloud image. A clutter score is estimated so that the clutter score describes extraneous features in the second 3D point cloud image. An overall score is computed as a difference between the coverage score and the clutter score.

Abstract: A system and method for three dimensional (3D) vision inspection using a 3D vision system. The system and method comprising acquiring at least one 3D image of a 3D object using the 3D vision system, using the 3D vision system; extracting a 3D visible runtime mask of the 3D image; using the 3D vision system, comparing the 3D runtime visible mask to a 3D reference visible mask; and, using the 3D vision system, determining if a difference of pixels exists between the 3D runtime visible mask and the 3D reference visible mask.

Abstract: This invention provides a system and method for the high-accuracy measurement of an object's surface displacement at a plurality of measurement points using a laser displacement sensor and a sensor process that corrects for noise and other accuracy-reducing factors. A camera assembly with an imager and optics package acquires images of an object surface, and a laser assembly projects a line on the object surface. These tasks can include (a) defining patches of the surface based upon a grid; (b) registering the object and aligning the grid with respect to the object; (c) excluding from the analysis of the image, irregular surface features; (d) oversampling one or more patches to improve overall accuracy; (e) specifying measurement resolution (i.e. in the (x, y z) dimensions), including choosing appropriate pixel sizes and number of measurements per measurement point; and (f) selecting optimal acquisition/image formation parameters.

Abstract: The present application discloses a probe placement module for placing probes on a virtual object depicted in an image. The probe placement module is configured to place probes on interest points of an image so that the probes can accurately represent a pattern depicted in the image. The probe placement module can be configured to place the probes so that the probes can extract balanced information on all degrees of freedom associated with the pattern's movement, which improves the accuracy of the model generated from the probes.

Abstract: This invention provides a system and method for determining the level of clutter in an image in a manner that is rapid, and that allows a scoring process to quickly determine whether an image is above or below an acceptable level of clutter—for example to determine if the underlying imaged runtime object surface is defective without need to perform a more in-depth analysis of the features of the image. The system and method employs clutter test points that are associated with regions on the image that should contain a low gradient magnitude, indicative of emptiness. This enables the runtime image to be analyzed quickly by mapping trained clutter test points at locations in the coordinate space in which lack of emptiness indicates clutter, and if detected, can rapidly indicate differences and/or defects that allow for the subject of the image to be accepted or rejected without further image analysis.

Abstract: A system and method for training multiple pattern recognition and registration models commences with a first pattern model. The model is trained from multiple images. Composite models can be used to improve robustness or model small differences in appearance of a target region. Composite models combine data from noisy training images showing instances of underlying patterns to build a single model. A pattern recognition and registration model is generated that spans the entire range of appearances of the target pattern in the set of training images. The set of pattern models can be implemented as either separate instances of pattern finding models or as a pattern multi-model. The underlying models can be standard pattern finding models or pattern finding composite models, or a combination of both.

Abstract: The disclosure relates to weighing moving objects in a weighing platform functionally coupled to a computer-vision tracking platform. The objects can translate, rotate, and translate and rotate. Weighing of the objects can be accomplished through combination of object imaging and upstream weighing. Object imaging can permit tracking, through computer vision, a logical object moving in a trajectory from the first location to the second location, wherein a logical object is a formal representation of one or more physical objects. Upstream weighing can permit updating a record indicative of weight of the one or more physical objects associated with the tracked logical object. As a part of weighing termination, data integrity check(s) can be performed on a plurality of records indicative of a weight of a single physical object. Based on outcome of the data integrity check(s), a record indicative of the weight of the single physical object can be supplied.

Abstract: This invention provides a system and method for determining correspondence between camera assemblies in a 3D vision system implementation having a plurality of cameras arranged at different orientations with respect to a scene involving microscopic and near microscopic objects under manufacture moved by a manipulator, so as to acquire contemporaneous images of a runtime object and determine the pose of the object for the purpose of guiding manipulator motion. At least one of the camera assemblies includes a non-perspective lens. The searched 2D object features of the acquired non-perspective image, corresponding to trained object features in the non-perspective camera assembly can be combined with the searched 2D object features in images of other camera assemblies, based on their trained object features to generate a set of 3D features and thereby determine a 3D pose of the object.

Abstract: A system and method for training multiple pattern recognition and registration models commences with a first pattern model. The model is trained from multiple images. Composite models can be used to improve robustness or model small differences in appearance of a target region. Composite models combine data from noisy training images showing instances of underlying patterns to build a single model. A pattern recognition and registration model is generated that spans the entire range of appearances of the target pattern in the set of training images. The set of pattern models can be implemented as either separate instances of pattern finding models or as a pattern multi-model. The underlying models can be standard pattern finding models or pattern finding composite models, or a combination of both.

Abstract: A method and apparatus for assessing at least one of motion linearity of a motion stage, stage motion straightness of a motion stage, image capture repeatability of a motion stage and camera and accuracy of a calibration plate used to assess motion stage characteristics, the method including using a line scan camera to generate two dimensional images of a calibration plate having a plurality imageable features thereon, examining the images to identify actual coordinates of the imageable features and using the actual coordinates to assess linearity, straightness, repeatability and/or plate accuracy.

Abstract: This invention provides a system and method for determining correspondence between camera assemblies in a 3D vision system implementation having a plurality of cameras arranged at different orientations with respect to a scene involving microscopic and near microscopic objects under manufacture moved by a manipulator, so as to acquire contemporaneous images of a runtime object and determine the pose of the object for the purpose of guiding manipulator motion. At least one of the camera assemblies includes a non-perspective lens. The searched 2D object features of the acquired non-perspective image, corresponding to trained object features in the non-perspective camera assembly can be combined with the searched 2D object features in images of other camera assemblies, based on their trained object features to generate a set of 3D features and thereby determine a 3D pose of the object.

Abstract: A method for training a pattern recognition algorithm including the steps of identifying the known location of the pattern that includes repeating elements within a fine resolution image, using the fine resolution image to train a model associated with the fine image, using the model to examine the fine image resolution image to generate a score space, examining the score space to identify a repeating pattern frequency, using a coarse image that is coarser than the finest image resolution image to train a model associated with the coarse image, using the model associated with the coarse image to examine the coarse image thereby generating a location error, comparing the location error to the repeating pattern frequency and determining if the coarse image resolution is suitable for locating the pattern within a fraction of one pitch of the repeating elements.

Abstract: An image of a semiconductor interconnection pad is analyzed to determine a geometric description of the zone regions of a multiple zone semiconductor interconnection pad. Edge detection machine vision tools are used to extract features in the image. The extracted features are analyzed to derive geometric descriptions of the zone regions of the pad, that are applied in semiconductor device inspection, fabrication, and assembly operations.

Abstract: A method for training a pattern recognition algorithm including the steps of identifying the known location of the pattern that includes repeating elements within a fine resolution image, using the fine resolution image to train a model associated with the fine image, using the model to examine the fine image resolution image to generate a score space, examining the score space to identify a repeating pattern frequency, using a coarse image that is coarser than the finest image resolution image to train a model associated with the coarse image, using the model associated with the coarse image to examine the coarse image thereby generating a location error, comparing the location error to the repeating pattern frequency and determining if the coarse image resolution is suitable for locating the pattern within a fraction of one pitch of the repeating elements.

Abstract: A method for training a pattern recognition algorithm for a machine vision system that uses models of a pattern to be located, the method comprising the steps of training each of a plurality of models using a different training image wherein each of the training images is a version of a single image of the pattern at a unique coarse image resolution, using the models to identify at least one robust image resolution where the image resolution is suitable for locating the pattern within an accuracy limit of the actual location of the pattern in the image and storing the at least one robust image resolution for use in subsequent pattern recognition processes.

Abstract: The disclosure relates to weighing moving objects in a weighing platform functionally coupled to a computer-vision tracking platform. The objects can translate, rotate, and translate and rotate. Weighing of the objects can be accomplished through combination of object imaging and upstream weighing. Object imaging can permit tracking, through computer vision, a logical object moving in a trajectory from the first location to the second location, wherein a logical object is a formal representation of one or more physical objects. Upstream weighing can permit updating a record indicative of weight of the one or more physical objects associated with the tracked logical object. As a part of weighing termination, data integrity check(s) can be performed on a plurality of records indicative of a weight of a single physical object. Based on outcome of the data integrity check(s), a record indicative of the weight of the single physical object can be supplied.

Abstract: A vision system is provided to determine a positional relationship between a semiconductor wafer on a platen and an element on a processing machine, such as a printing screen, on a remote side of the semiconductor wafer from the platen. A source directs ultraviolet light through an aperture in the platen to illuminate the semiconductor wafer and cast a shadow onto the element adjacent an edge of the semiconductor wafer. A video camera produces an image using light received from the platen aperture, wherein some of that received light was reflected by the wafer. The edge of the semiconductor wafer in the image is well defined by a dark/light transition.

Abstract: Digital image processing methods are applied to an image of a semiconductor interconnection pad to preprocess the image prior to an inspection or registration. An image of a semiconductor pads exhibiting spatial patterns from structure, texture or features are filtered without affecting features in the image not associated with structure or texture. The filtered image is inspected in a probe mark inspection operation.