Abstract: A method and an image processing system for detecting an object in an image are described. A set of line segments are detected in the image. A subset of the line segments is identified based on a projection space orientation that defines a projection space. Each one of the line segments of the subset of line segments is projected into the projection space to obtain a set of projected line segments, where each projected line segment of the set of projected line segments is represented by a respective set of projection parameters. A determination is performed, in the projection space, based on the sets of projection parameters and a shape criterion that characterizes the object, of whether the image includes an instance of the object. In response to determining that the image includes the instance of the object, the instance of the object is output.

Abstract: A method for calibrating a vehicle sensor of a motor vehicle. The method includes: ascertaining sensor data for a plurality of measuring points in time during a total measuring time period, the total measuring time period being subdivided into partial measuring time periods, and the motor vehicle moving relative to objects in surroundings of the motor vehicle; for each partial measuring time period, computing positions of the objects based on the ascertained sensor data; for each partial measuring time period, computing a partial measuring time period sinogram based on the computed positions; computing a total measuring time period sinogram by adding the partial measuring time period sinograms and correcting using a factor that is a function of the partial measuring time period sinograms; ascertaining an orientation of the vehicle sensor based on the total measuring time period sinogram; and calibrating the vehicle sensor based on the ascertained orientation.

Abstract: A computer-implemented method for sharing information about a subject without disclosing identification information about the subject includes comparing, by a first computer system, a third set of data with a fourth set of data. The third set of data is converted by a transformation from a first set of identification data of a first subject provided by the first computer system. The fourth set of data is converted by the transformation from a second set of identification data of a second subject provided by a second computer system. The transformation renders the first subject unidentifiable from the third set of data and the second subject unidentifiable from the fourth set of data. The computer-implemented method also includes transmitting, from the first computer system to the second computer, a fifth set of data when the third set of data corresponds to the fourth set of data.

Abstract: Vision-aided inertial navigation techniques are described. In one example, a vision-aided inertial navigation system (VINS) comprises an image source to produce image data at a first set of time instances along a trajectory within a three-dimensional (3D) environment, wherein the image data captures features within the 3D environment at each of the first time instances. An inertial measurement unit (IMU) to produce IMU data for the VINS along the trajectory at a second set of time instances that is misaligned with the first set of time instances, wherein the IMU data indicates a motion of the VINS along the trajectory. A processing unit comprising an estimator that processes the IMU data and the image data to compute state estimates for 3D poses of the IMU at each of the first set of time instances and 3D poses of the image source at each of the second set of time instances along the trajectory.

Abstract: The present application discloses a method for image text recognition, an apparatus, a device, and a storage medium, and relates to image processing technologies in the field of cloud computing. A specific implementation is: acquiring an image to be processed, where at least one text line exists in the image to be processed; processing each text line in the image to be processed to obtain a composite encoded vector corresponding to each word in each text line, where the composite encoded vector carries semantic information and position information; and determining a text recognition result of the image to be processed according to the semantic information and the position information carried in the composite encoded vector corresponding to each word in each text line. This technical solution can accurately distinguish adjacent fields with small pixel spacing in the image and improve the accuracy of text recognition in the image.

Abstract: An information detection method includes: determining key point information in a target identification from a target image based on a preset deep learning algorithm; obtaining an image of the target identification from the target image according to the key point information; and determining information of a preset field from the image of the target identification according to the image of the target identification and a preset identification template matching the target identification.

Abstract: A method for detecting faults in substring search operations includes providing, using a processor unit including vector registers of M vector elements each, an M×M matrix of comparators for characterwise comparison of the elements of a reference string stored in a first one of the vector registers and a target string stored in a second one of the vector registers. A vector element is an n-bit element for encoding a character. A resulting bit vector is generated using comparison performed by the M×M matrix. The resulting bit vector indicates characters of the target string that fully match the reference string and indicates characters of the target string that partially match the reference string. Fault detection in the substring search operations is performed by utilizing the resulting bit vector.

Abstract: One embodiment of the present invention sets forth a technique for analyzing transcriptions of a recording. The technique includes storing per-character differences between a first set of characters from a first transcription of the recording and a second set of characters from a second transcription of the recording in a matrix with a fixed width. The technique also includes encoding the per-character differences in the matrix into a vector of the fixed width. The technique further includes outputting the vector as a representation of a pairwise error rate between the first transcription and the second transcription.

Abstract: A method for determining position of a target includes determining at least two measurement points of the target, obtaining vector data of at least two direction vectors based on an imaging position of the target on an imaging system and parameter data of the imaging system, and determining a distance between the target and the imaging system according to the vector data and a current vertical height of the imaging system. Each of the at least two direction vectors is a vector from the imaging system to a corresponding one of the at least two measurement points.

Abstract: Techniques for time-correlated ink are described. According to various embodiments, ink input is correlated to content. For instance, ink input received during playback of a video is timestamped. According to various embodiments, ink input displayed over content is removed after input ceases. Further, ink input is displayed during playback of the portion of content to which the ink input is time correlated.

Abstract: Systems and methods for performing OCR of an image depicting text symbols. An example method comprises: receiving an original image of a document; identifying location of a symbol sequence in the image; performing a series of conversion operations on a the portion of the image containing the identified symbol sequence; performing OCR of the symbol sequence in the resulting images; and combining these OCR results to produce a resulting OCR text representing at least a portion of the original document.

Abstract: A processor is described having a functional unit of an instruction execution pipeline. The functional unit has comparison bank circuitry and adder circuitry. The comparison bank circuitry is to compare one or more elements of a first input vector against an element of a second input vector. The adder circuitry is coupled to the comparison bank circuitry to add the number of elements of the second input vector that match a value of the first input vector on an element by element basis of the first input vector.

Abstract: Methods, apparatuses, and embodiments related to aligning an image of a table of a form with an image of a table of a template of the form. Automatically extracting data entered in fields of a table of a form by a user can be aided by matching the form with a template of the form. The form template can have a digitized representation that identifies locations of fields of the form, and that identifies labels of the fields. Matching the form with the form template can enable locations and labels of fields of the form to be identified based on the digitized representation. However, matching the form with the form template may require matching the table of the form with the table of the form template, and matching two tables can be challenging. For example, the tables can be rotated, warped, scaled, etc. relative to each other.

Abstract: A vehicle includes a camera capturing an image of a road sign associated with a road that the vehicle is traveling on. A first processing device is communicatively coupled to the camera and recognizes text characters in the captured image. A second processing device is communicatively coupled to the first processing device and converts the recognized text characters into an electronic speech signal. A loudspeaker is communicatively coupled to the second processing device and converts the electronic speech signal into audible speech that can be heard by the driver of the vehicle.

Abstract: A method and apparatus for associating strokes with content are described. In one embodiment, an apparatus comprises: a memory to store contexts for a plurality of strokes, wherein each context represents an association between at least one stroke and a portion of image data previously displayed; and a matching module coupled to the memory to: receive an indication of a new stroke to be associated a first image and a first context, determine if the portion the first context is the same as, overlaps or is independent of the contexts stored in the memory, and update the memory to include the new stroke by adding the new stroke to one of the contexts stored in the memory or storing the new stroke with the first context independent of the contexts already stored in the memory.

Abstract: A coordinate-conversion-parameter determination apparatus includes a marker detector that detects feature points of a marker from each of a first and second images; a three-dimensional position determiner that calculates three-dimensional coordinates of each of the feature points of the marker in a camera coordinate system in which reference is a first place, based on positions of the feature points in the first and second images; and a coordinate-conversion-parameter determiner that calculates a coordinate conversion parameter so that a sum of absolute values of errors between three-dimensional coordinates in a world coordinate system, the three-dimensional coordinates in the world coordinate system being obtained by converting the three-dimensional coordinates of the feature points of the marker in the camera coordinate system by using the coordinate conversion parameter, and three-dimensional coordinates of feature points in the world coordinate system that correspond to the feature points of the marker in

Abstract: The present disclosure relates to a gaze based error recognition detection system that is intended to predict intention of the user to correct user drawn sketch misrecognitions through a multimodal computer based intelligent user interface. The present disclosure more particularly relates to a gaze based error recognition system comprising at least one computer, an eye tracker to capture natural eye gaze behavior during sketch based interaction, an interaction surface and a sketch based interface providing interpretation of user drawn sketches.

Abstract: Techniques are disclosed for facilitating optical character recognition (OCR) by identifying one or more regions in an electronic document to perform the OCR. For example a method for identifying information in an electronic document includes obtaining a set of training documents for each template of a plurality of templates for the electronic document, extracting spatial attributes for at least a first label region and at least a first corresponding value region from the set, and training a classifier model based on the extracted spatial attributes, wherein the classifier model is used to identify the information in the electronic document. The spatial attributes represent a position of at least the first label region and at least the first value region within the electronic document.

Abstract: Font recognition and similarity determination techniques and systems are described. In a first example, localization techniques are described to train a model using machine learning (e.g., a convolutional neural network) using training images. The model is then used to localize text in a subsequently received image, and may do so automatically and without user intervention, e.g., without specifying any of the edges of a bounding box. In a second example, a deep neural network is directly learned as an embedding function of a model that is usable to determine font similarity. In a third example, techniques are described that leverage attributes described in metadata associated with fonts as part of font recognition and similarity determinations.

Abstract: A method for normalizing at least one character included in an image is provided. The method includes steps of: (a) an apparatus for normalizing a specific character adding or supporting to add at least one margin, which is determined by referring to information on at least one another character related to the specific character detected in the image, around a bounding box which includes the specific character; and (b) the apparatus normalizing or supporting to normalize an area of the bounding box around which the at least one margin is added.

Abstract: An object detection apparatus for detecting a target object in an input image. The apparatus includes a storage storing, for each of a plurality of part areas forming an area subject to image recognition processing, image recognition dictionaries used to recognize a target object and typed according to variations in appearance of a part of the target object to be detected in the part area. A part score calculator calculates, for each of the part areas, a part score indicative of a degree of similarity between the part area and each of at least some of the image recognition dictionaries. An integrated score calculator calculates an integrated score that is a weighted sum of the part scores for the respective part areas. A determiner determines, on the basis of the integrated score, whether or not the target object is present in the subject area.

Abstract: A vertical and horizontal line detection method for document images includes generating multiple binary images from the input grayscale document image based on multiple binarization thresholds, detecting horizontal and vertical lines in each of the multiple binary images independently, and merging the detection results from the multiple binary images. The line detection process for each binary image include applying an opening operation using a vertical or horizontal line as the structuring element, and removing connected components that are not vertical or horizontal lines based on a stroke width analysis. The boundaries of the detected lines are obtained using horizontal and vertical projections.

Abstract: A system and method for providing rewards to a user in a virtual space based on the user performance of gestures is disclosed. Gesture parameters may be obtained for physical actions performed by virtual space users in the real world. Gestures and corresponding gesture commands in the virtual space may be identified based on the gesture parameters. User performance of an identified gesture may be qualified by one or more gesture metrics. The gesture metrics may be determined based on the obtained gesture parameters. Access to reward content may be provided to the user based on user performance of the identified gestures as qualified by the determined gesture metrics.

Abstract: The invention provides an electronic apparatus. In one embodiment, the electronic apparatus is coupled to a display device, and comprises a memory, a data transmission interface, and a control module. The memory comprises a virtual frame buffer for storing data to be transmitted to the display device. The data transmission interface performs data transmission between the electronic apparatus and the display device. The control module stores a current character image corresponding to a current character in the virtual frame buffer when users input the current character to an input device, determines whether a prior image which has not been transmitted to the display device exists in the virtual frame buffer, determines whether the current character image matches the prior image, and combines the prior image with the current character image to obtain a combined image in place of the prior image when the current character image matches the prior image.

Abstract: This invention relates to a system that adaptively compensates for subject motion in real-time in an imaging system. An object orientation marker (30), preferably a retro-grate reflector (RGR), is placed on the head or other body organ of interest of a patient (P) during a scan, such as an MRI scan. The marker (30) makes it possible to measure the six degrees of freedom (x, y, and z-translations, and pitch, yaw, and roll), or “pose”, required to track motion of the organ of interest. A detector, preferably a camera (40), observes the marker (30) and continuously extracts its pose. The pose from the camera (40) is sent to the scanner (120) via an RGR processing computer (50) and a scanner control and processing computer (100), allowing for continuous correction of scan planes and position (in real-time) for motion of the patient (P).

Abstract: A method, non-transitory computer readable medium and character recognition device for character recognition includes a character recognition device for receiving an image representing a character comprising one or more strokes. A set of first parameters associated with each of the one or more first strokes is determined. Each of the one or more first strokes is compared with the associated set of first parameters with a plurality of stored sets of second parameters, where each of the plurality of stored second strokes is associated with a stored set of second parameters. A second stroke is identified from among the plurality of stored second strokes corresponding to each of the one or more first strokes based on the comparison. The character is identified based on the identified one or more second strokes.

Abstract: A method is provided for identifying one or more scored candidate objects that may correspond to one or more actual pallets in a gray scale image. The method may comprise: identifying, by a computer, a first plurality of scored candidate objects in the gray scale image; storing, by the computer, a list of the first plurality of scored candidate objects, wherein the list includes a respective record for each of the first plurality of scored candidate objects; determining, by the computer, a subset of the first plurality of scored candidate objects to eliminate from the list based on a comparison amongst the respective records of the first plurality of scored candidate objects; and removing, by the computer, the subset of the first plurality of scored candidate objects from the list to create an updated list for a second plurality of scored candidate objects, wherein the updated list includes a respective record for each of the second plurality of scored candidate objects.

Abstract: Systems and methods for synthesizing paintbrush strokes may use high fidelity pose data of reference strokes to supplement lower dimension stroke data. For example, 6DOF data representing reference strokes created by skilled artists may be captured and stored in a library of reference strokes. A query stroke made by a less skilled artist/user and/or made using input devices that do not provide 6DOF data may be processed by a stroke synthesis engine to produce an output stroke that follows the trajectory of the query stroke and includes pose data from one or more reference strokes. The stroke synthesis engine may construct feature vectors for samples of reference strokes and query strokes, select the best neighbor feature vector from the library for each query stroke sample, assign the pose data of the best neighbor to the query sample, and smooth the sequence of assigned poses to produce the synthesized output stroke.

Abstract: A first technique of recognizing content is disclosed, including: determining a first value representative of a pixel content present at a first set of pixels associated with a first distance from a pixel under consideration; determining a second value representative of a pixel content present at a second set of pixels associated with a second distance from the pixel under consideration; and using the first and second values to compute one or more spatial features associated with the pixel under consideration for purposes of content recognition. A second technique of recognizing content is also disclosed, including: determining, for a pixel, a first value representative of a first feature associated with a set of pixels associated with a first direction from the pixel; and determining, for the pixel, a second value representative of a second feature associated with a set of pixels associated with a second direction from the pixel.

Abstract: A handwritten character input device includes a correction amount computation unit which performs, when a characteristic value of handwritten characters input to a designated character writing area of a template image by a handwriting input operation of a user does not meet predetermined character writing requirements, a computation process that computes a correction amount of the characteristic value required to meet the character writing requirements; and a corrected handwritten character information output unit which outputs information of after-correction handwritten characters generated based on the correction amount computed by the correction amount computation unit, the characteristic value of the after-correction handwritten characters meeting the character writing requirements.

Abstract: Integrating features is disclosed, including: determining a value associated with a temporal feature for a point; determining a value associated with a spatial feature associated with the temporal feature; including the value associated with a spatial feature and the value associated with the temporal feature into a feature vector; and using the feature vector to decode for a character. Determining a transform is also disclosed, including: determining, for a point associated with a sequence of points, a set of points including: the point, a first subset of points of the sequence preceding a sequence position associated with the point, and a second subset of points following the sequence position associated with the point; and determining the transform associated with the point based at least in part on the set of points.

Abstract: A linear mark detection system includes a filter value calculation unit, a filter value threshold processing unit, and a feature value output unit. The filter value calculation unit calculates, for each pixel of an image, filter values corresponding to each of a plurality of linear mark widths using pixel values inside and outside the linear mark. The filter value threshold processing unit selects the most suitable value from filter values corresponding to each of the plurality of linear mark widths as a feature value. The feature value output unit outputs, for each pixel of the image, the feature value computed by the filter value threshold processing function as the feature value corresponding to the linear mark.

Abstract: A method is provided for identifying one or more scored candidate objects that may correspond to one or more actual pallets in a gray scale image.

Abstract: A method is provided for evaluating a possible center stringer of a pallet. The method may comprise providing a first Ro image; providing a second Ro image comprising pixels that may generally correspond to an orthogonal distance from an origin point to one or more possible vertical right lines in the corresponding gray scale image; identifying, using a computer, a possible lower left corner location of a center stringer in the corresponding gray scale image; determining, using the computer, an upper left corner location based on the possible lower left corner location; identifying, using the computer, a possible lower right corner location of the center stringer in the corresponding gray scale image; and determining, using the computer, an upper right corner location based on the possible lower right corner location.

Abstract: The recognition of a lane boundary line expressed by a stud-shaped lane mark is realized with high accuracy against noise and with a small calculation amount and a small memory consumption amount. A lane recognition device extracts a candidate position of a lane mark from a received input image. A vote value is weighted according to the elapsed time with regard to the extracted candidate position of the lane mark to cumulatively perform voting to a parameter space of an approximate curved line or an approximate straight line, and, thus, to generate a time-series cumulative voting feature image. A candidate of a lane boundary line position is extracted based on the generated time-series cumulative voting feature image to determine the lane boundary line position. The lane recognition device may be provided with an image output device which images an image and outputs the imaged image.

Abstract: Methods for correcting distortions in an image including text, or an image of a page that includes text, are disclosed. The methods include identifying reliable and substantially straight lines from elements in the image. Vanishing points are determined from the lines. Parameters associated with a rectangle are determined. A coordinate conversion is performed.

Abstract: Methods, systems, and articles of manufacture for the automatic determining of calls to action in a webpage are disclosed. These include, rendering a web page in a browser, capturing the rendered web page as an image, filtering the image using one or more image filters, and identifying the calls to action in the filtered image using at least one call to action selection criteria.

Abstract: A technique that uses repetitive and reliably recognizable parts of handwriting, during digital handwriting data entry, to trigger recognition of digital ink and to repurpose handwriting task area properties. In one example embodiment, this is achieved by drawing one or more delayed strokes of a desired sub-word unit using a stylus on a touch screen. An associated data of the drawn one or more strokes is inputted via the touch screen into a handwriting recognition engine. A first trigger stroke in the drawn one or more strokes that can be used to trigger the sub-word unit recognition by the handwriting recognition engine is then determined. The sub-word unit recognition is then triggered for the drawn one or more strokes based on the determined first trigger stroke by the handwriting recognition engine.

Abstract: A computer-implemented method can comprise accessing a plurality of pixels representing an image and identifying at least two scanlines in the plurality of pixels. By analyzing the scanlines, a computing device carrying out the method can determine if the image is suited for slicing and, if the image is suited for slicing, the device can determine a slicing strategy by analyzing pixel values of the at least two scanlines. Data indicating the slicing strategy can be used to carry out a resizing operation and/or to generate structured code based on the slicing strategy, such as HTML and CSS code to generate a resizable element corresponding to the image. The slicing strategy can be determined independent of input defining or adjusting boundaries between slices.

Abstract: An electronic writing solution server includes a query engine and a user interface engine. The query engine receives a search query from a user, the search query specifying a field in at least one form and including a first stroke or text, retrieves clusters from a database, retrieves example strokes from the clusters at vary levels of similarity to the first stroke, receives user selections for strokes of interest from the example strokes, and retrieves information associated with the strokes of interest from the database. The user interface engine generates graphical data for presenting the retrieved information to the user.

Abstract: A display panel test apparatus includes: an image pickup part which picks up an image from a target display panel; a jig including a receiving part which receives the target display panel, a fixing part which fixes the image pickup part, and an adjusting part which adjusts an image pickup angle of the image pickup part; a pattern generating part which provides the target display panel with a test pattern; a defect extracting part which analyzes test image data provided from the image pickup part using a defect extracting algorithm and extracts display defect information, where the defect extracting algorithm includes different settings corresponding to different types of display defects; and a control part which generates evaluated data corresponding to a viewing angle of the target display panel using the image pickup angle of the image pickup part and the display defect information.

Abstract: In a first exemplary embodiment of the present invention, an automated, computerized method is provided for processing an image. According to a feature of the present invention, the method comprises the steps of providing an image file depicting an image, in a computer memory, organizing information relevant to the image according to a smooth-illumination constraint enforcing a Laplacian filter response such that ?2ƒ(x, y)=0 and utilizing the smooth-illumination constraint to generate an intrinsic image from the image.

Abstract: Methods and systems of recognizing images may include an apparatus having a hardware module with logic to, for a plurality of vectors in an image, determine a first intermediate computation based on even pixels of an image vector, and determine a second intermediate computation based on odd pixels of an image vector. The logic can also combine the first and second intermediate computations into a Hessian matrix computation.

Abstract: An image processing apparatus includes a storage unit, a receiving unit, a search unit, a processing unit and an output unit. The storage unit stores information of drawn image elements. The receiving unit receives a transformation instruction to transform a part of the image elements specified by the information of the image elements. The search unit searches the storage unit for an image element that (i) has a relation, being based on a predetermined condition, with the image elements, which are a target of the transformation instruction, and (ii) is not the target of the transformation instruction. The processing unit performs a transformation process on (i) the image elements, which are the target of the received transformation instruction, and (ii) the image element, which is found as the search result, based on the received transformation instruction. The output unit outputs the result of the transformation process.

Abstract: Target detecting method and apparatus are disclosed. In the target detecting method, edges in a first direction in an input image may be detected to obtain an edge image comprising a plurality of edges in the first direction; and one or more candidate targets may be generated according to the plurality of edges in the first direction, a region between any two of the plurality of edges in the first direction in the input image corresponding to one of the candidate targets.

Abstract: To detect a straight line using the Hough transform taking into consideration not only the number of points but also other properties of the straight line, the Hough transform unit performs a Hough transform on contour-enhanced binary image data. The Hough table stores a count after the Hough transform. The adjustment unit adjusts the count. The straight line detection unit detects a straight line based on the adjusted count. Additionally, to detect a straight line, independent of its direction or location in the image, for each straight line in the binary image data, the straight line calculation unit determines the intersections where that straight line cuts up the binary image data to calculate the intersection distance. The normalization unit divides the count stored in the Hough table by the intersection distance to normalize the count. The straight line detection unit detects a straight line based on the normalized count.

Abstract: A system with for capturing and storing an image of a user's handwritten alphanumeric characters. The characters are stored in memory in association with a standard alphanumeric character set such as an ASCII character set. A text document can then be transcribed into the stored handwritten alphanumeric characters for resembling a handwriting of the user. The characters can be represented by mathematical functions such as Taylor Series or McLauren Series and can be digitally reproduced based on the stored functions.

Abstract: When photographing images including a person's face at an event, photographers tend to photograph images so that features of the event appear in a region around the person's face. An image data processing device of the present invention extracts image feature information so that an image feature calculated based on pixels in the region around the person's face, which tends to represent features of an event, is reflected more than that calculated based on pixels in a region remote from the person's face, which tends not to represent features of an event. This allows the image data processing device to calculate image feature information reflecting features of an event more than that calculated by a conventional image data processing device. The image data processing device therefore improves classification precision compared to the conventional device when classifying images using the image feature information calculated by the image data processing device.

Abstract: A document recognition system and method, where images are represented as a collection of primitive features whose spatial relations are represented as a graph. Useful subsets of all the possible subgraphs representing different portions of images are represented over a corpus of many images. The data structure is a lattice of subgraphs, and algorithms are provided means to build and use the graph lattice efficiently and effectively.

Abstract: A method is provided for tracing a line from a possible corner of a pallet. The method may comprise: providing a Ro image; identifying, using a computer, a possible pallet corner in the Ro image; positioning, using the computer, a J×K window over at least respective portions of a plurality of rows in the Ro image including at least a portion of a row containing the possible corner; calculating, using the computer, an average of pixel values for each row in the J×K window; determining, using the computer, one of the rows in the J×K window having an average pixel value nearest a current pixel location being considered for inclusion in a line being traced and defining the one row as a nearest row; and deciding, using the computer, whether the nearest row is over a pallet.