Provided is a contour shape recognition method, including: sampling and extracting salient feature points of a contour of a shape sample; calculating a feature function of the shape sample at a semi-global scale by using three types of shape descriptors; dividing the scale with a single pixel as a spacing to acquire a shape feature function in a full-scale space; storing feature function values at various scales into a matrix to acquire three types of feature grayscale map representations of the shape sample in the full-scale space; synthesizing the three types of grayscale map representations of the shape sample, as three channels of RGB, into a color feature representation image; constructing a two-stream convolutional neural network by taking the shape sample and the feature representation image as inputs at the same time; and training the two-stream convolutional neural network, and inputting a test sample into a trained network model to achieve shape classification.

A standard image of a product to be a standard for an inspection target is displayed, to set a first region so as to surround a standard pattern in the standard image. Further, a second region for characterizing a position and a posture of the standard pattern is set in the standard image. In a first search step, a feature extracted from the first region set in the standard image is searched from an inspection target image, to roughly obtain the position and the posture of the standard pattern in the inspection target image. In the second search step, the feature extracted from the second region set in the standard image is searched from the inspection target image, to minutely obtain at least one of the position and the posture of the standard pattern in the inspection target image.

A method for analyzing and searching 3D models includes steps of obtaining data global features and data local features of data images by globally analyzing and locally analyzing data images of 3D models respectively; obtaining searching global features and searching local features by globally analyzing and locally analyzing searching images respectively; obtaining corresponding data global features and corresponding data local features based on the search global features and the searching local feature; and obtaining corresponding data images based on the corresponding data global features and the corresponding data local features.

A positioning apparatus includes: a calculation unit that calculates an amount of deviation between a position of a feature point of a reference workpiece and a feature point of a workpiece by comparing a relative position of an imaging unit with respect to a table when the workpiece is imaged by the imaging unit with a reference relative position, and comparing a position of a feature point of the workpiece in the image of the workpiece imaged by the imaging unit with a reference point image position; and a program changing unit that generates a correction amount such that the amount of deviation calculated by the calculation unit becomes zero, and thereby changes a program of the machine tool.

In one aspect, a computer-implemented method for conflating a base dataset with a secondary dataset may generally include defining a locker boundary around each of a plurality of base polylines of the base dataset and identifying a plurality of initial matched segments and a plurality of initial mismatched segments for a plurality of secondary polylines of the secondary dataset, wherein each portion of the secondary polylines that is included within a locker boundary is defined as an initial matched segment and each portion of the secondary polylines that is not included within a locker boundary is defined as an initial mismatched segment. The method may also include identifying an offset parameter defined between a first initial matched segment and its corresponding base polyline using a three-vertex approximation and, if the offset parameter exceeds a predetermined offset threshold, defining the first initial matched segment as a mismatched segment.

Disclosed is an image processing device that includes: a storage unit that holds model information including information indicating a plurality of feature points of a detection object; a correspondence-relationship determination unit that extracts a plurality of feature points included in an input image as a two-dimensional image or a three-dimensional image and that determines a correspondence relationship between an extracted plurality of feature points and a plurality of feature points of the model information; and a position estimating unit that estimates, based on the correspondence relationship, one or a plurality of second positions of the input image corresponding to one or a plurality of first positions set relatively to the detection object. The one or plurality of first positions is included in a point, a line, or a region.

Methods and systems for diagnosing semiconductor wafer are provided. A target image is obtained according to graphic data system (GDS) information of a specific layout in the semiconductor wafer, wherein the target image includes a first contour having a first pattern corresponding to the specific layout. Image-based alignment is performed to capture a raw image from the semiconductor wafer according to the first contour. The semiconductor wafer is analyzed by measuring the raw image, so as to provide a diagnostic result.

A shape similarity metric can be provided that indicates how similar two or more shapes are. A difference between a union of the shapes and an intersection of the shapes can be used to determine the similarity metric. The shape similarity metric can provide an average distance between the shapes. Different processes for determining shapes can be evaluated for accuracy based on the shape similarity metric. New or alternative shape-determining processes can be compared for accuracy against other shape-determining processes including reference shape-determining processes. Shape similarity metrics can be determined for two-dimensional shapes and three-dimensional shapes.

The present disclosure provides systems and methods to enable detection of linear objects such as utility poles in complex and heterogeneous outdoor surroundings. The methods deal with shape and orientation as prominent features of a pole model. Candidate trapeziums from 2D images of the face of the poles are extracted, some of which represent parts of the pole. To overcome the missed detection of certain parts due to problems of occlusion and diffusion into background, shape based techniques, that extrapolate and capture a longer trapezium representing the pole is implemented. The region growing stage or extrapolation is driven by orientation-based clustering of trapeziums. Context information is further used to identify objects of interest, by discarding false positives. Besides detecting poles of interest, the detected poles are further analysed to identify damages, if any.

A method of labelling a water surface within an image is provided. The method includes: receiving image data of the image generated by a camera, the image including at least one water surface; receiving water surface extension data, the water surface extension data being representative of an area over which the water surface extends in the real world; matching the water surface extension data to the image data based on a spatial relationship between the camera and the area of the water surface; and labelling the water surface in the image based on the matched water surface extension data.