74 results about "Horizontal line segment" patented technology

The invention discloses a method and a device for recognizing table cells in a scanned image, comprising the following steps: acquiring horizontal line segments and vertical line segments in the scanned image of a table document; removing the horizontal line segments and the vertical segments which are less than the first threshold value in the scanned image, wherein, the first threshold value isset according to the height of the minimum character in the scanned image and the resolution ratio of the scanned image; and recognizing the table cells in the scanned image according to the remainedhorizontal line segments and vertical line segments. The invention not only has the characteristic of high recognition success rate of the traditional line detection algorithm, but also can improve the recognition speed of the table cells in the scanned image under the premise of ensuring high recognition rate.

This invention provides a correcting device and a correcting method for perspective transformation of document images. The correcting device comprises a horizontal vanishing point determining unit, for detecting a horizontal vanishing point of the perspective transformed document image; a vertical vanishing point determining unit, for detecting a vertical vanishing point of the perspective transformed document image; and a perspective transformation correcting and converting unit, for correcting the perspective transformed document image; wherein the horizontal vanishing point determining unit comprises a direct horizontal line segment detecting unit, an indirect horizontal line segment detecting unit and a horizontal vanishing point detecting unit, and wherein the horizontal vanishing point detecting unit detects a horizontal vanishing point in accordance with a direct horizontal line segment detected by the direct horizontal line segment detecting unit and an indirect horizontal line segment detected by the indirect horizontal line segment detecting unit.

A computerized method of recognizing an input hand-drawn table formed by freeform line objects comprises transforming each freeform line object of the table into one of a vertical line segment and a horizontal line segment, generating a grid system based on the vertical and horizontal line segments and converting the generated grid system into a table object.

A computerized method of recognizing an input hand-drawn table formed by freeform line objects comprises transforming each freeform line object of the table into one of a vertical line segment and a horizontal line segment, generating a grid system based on the vertical and horizontal line segments and converting the generated grid system into a table object.

The invention provides a vanishing point detecting device and a detecting method, and the vanishing point detecting device comprises a line segment detecting unit used for obtaining a horizontal line segment or a vertical line segment of an input image, an intersection point set obtaining device used for obtaining the intersection point set of the horizontal line segment or an intersection point set of the vertical line segment, a candidate vanishing point detecting unit used for determining a candidate vanishing point, and a vanishing point determining unit used for determining a final vanishing point according to the candidate vanishing point determined by the candidate vanishing point detecting unit; wherein the candidate vanishing point detecting unit is one or more of the following: a candidate vanishing point detecting unit based on fanning beam projection and density estimation, which carries out fanning beam projection to the input image and takes the intersection point with the highest point density in a high-density fanning beam area as the candidate vanishing point; a candidate vanishing point detecting unit based on fanning beam projection and clustering, which carries out the fanning beam projection, obtains a plurality of clustering of the intersection points in the high-density fanning beam area and takes the center of each clustering as the candidate vanishing point.

The invention discloses a container contour positioning method based on edge detection. The container contour positioning method adopts a camera to collect images on two opposite side of the container and obtains a rough positioning range for an upper lock hole and a lower lock hole; lock hole rough positioning images are converted to gray scale images to perform pre-processing to obtain images of I1,I2; edge detection is performed on the image I1 and the image I2 to obtain edge images I1', I'2 of the upper lock hole and the lower lock hole which are roughly positioned; a Hough straight line detection algorithm is adopted to perform detection to obtain a line segment set; a DBCAN clustering algorithm is adopted to maintain line segments close to horizontal and vertical to obtain a horizontal line segment set LH and a vertical line segment set LV; four intersection points p1, p2, p3, and p4 are obtained and the four intersection points constitutes a quadrangle which is the couture of the container. As a result, the positioning of the container couture is realized. The container contour positioning method based on the edge detection can effectively solve a problem of an interference edge of the lock hole image during the edge detection and can overcome a defect that the Hough straight line effect is not good.

This invention provides a correcting device and a correcting method for perspective transformation of document images. The correcting device comprises a horizontal vanishing point determining unit, for detecting a horizontal vanishing point of the perspective transformed document image; a vertical vanishing point determining unit, for detecting a vertical vanishing point of the perspective transformed document image; and a perspective transformation correcting and converting unit, for correcting the perspective transformed document image; wherein the horizontal vanishing point determining unit comprises a direct horizontal line segment detecting unit, an indirect horizontal line segment detecting unit and a horizontal vanishing point detecting unit, and wherein the horizontal vanishing point detecting unit detects a horizontal vanishing point in accordance with a direct horizontal line segment detected by the direct horizontal line segment detecting unit and an indirect horizontal line segment detected by the indirect horizontal line segment detecting unit.

What is disclosed is a system and method for determining whether a front seat in a motor vehicle is occupied based on seat pattern recognition. The present invention takes advantage of the observation that an unoccupied seat of a motor vehicle exhibits features which are distinguishable from an occupied seat. An unoccupied motor vehicle seat typically features long contiguous horizontal line segments and curve segments, and substantially uniform areas encompassed by these segments which are not present in an occupied seat. The present method provides a long horizontal edge test which uses location information within a defined window of the image, edge linking, softness of the edge, number of lines, line / curve fitting, and other techniques to locate horizontal edges in the image which define a seat, and a uniformity step which determines whether the area bounded by the horizontal edges is relatively uniform indicating an unoccupied seat.

The embodiment of the invention discloses a method, device and equipment for carrying out three-dimensional reconstruction on a two-dimensional image and a storage medium, and belongs to the technicalfield of image processing. The method comprises steps of obtaining a two-dimensional picture set; performing feature point extraction on a single picture in the two-dimensional picture set to obtaina feature point set; obtaining the size of the picture and the position information of the feature points in the feature point set in the corresponding picture; acquiring position information of the feature points in a world coordinate system based on an OpenCV camera calibration method; based on an LSD straight line detection method and the position information of the feature points in the worldcoordinate system, obtaining a horizontal line segment in the picture; constructing a three-dimensional grid chart corresponding to the picture based on a constraint Delaunay algorithm; and combiningthe three-dimensional grid diagrams constructed by the pictures in the two-dimensional picture set to obtain a three-dimensional surface model of the object, thereby completing three-dimensional reconstruction. According to the invention, three-dimensional simulation reconstruction can be carried out on the two-dimensional picture rapidly and accurately; and the manpower and material resource consumption during three-dimensional reconstruction can be reduced.

An image processing method is capable of generating an image containing graphics that are transformed without carrying out multiplying / dividing operations. Image data of a transformed graphic is generated from image data of a source graphic stored in an image memory, and also stored into a frame buffer. Coordinate values of points of the graphic being transformed are determined utilizing digital differential analysis (DDA) algorithm-based processes. Image data of a location specified by the horizontal and vertical coordinate values of each point on the source graphic is read out from the image memory, and written into the frame buffer as image data of each point on the horizontal line segment.

The invention discloses a data increment updating method. The method comprises the following steps: constructing a similar information set, constructing a horizontal line segment graph, constructing a path graph, constructing a minimum cost path, constructing an increment packet and applying the increment packet. With the adoption of the technical scheme, the similar degree between new and old files is represented in the form of a graph, the problem of generating a minimum increment packet is converted into the problem of finding a shortest path, and the minimum increment packet is generated according to the path; the data increment updating is carried out based on the method disclosed by the invention and the data quantity can be averagely reduced by 69.3%; compared with the data increment updating method in the prior art, the compression ratio is high, and the operation time of applying the increment packet is shorter; and the method has a wide application range, is not only applicable to consumer electronics, but also can be applied to other platforms and systems.

The invention discloses a machine vision-based tire tread detection method and provides a machine vision-based tire tread detection method. According to the technical scheme of the invention, based on the machine vision software halcon technology and the image processing algorithm, the system reliability and the measurement accuracy are ensured, and the production efficiency and the product quality are improved. In this way, the intelligent tire tread detection is realized and the urgent demand of social production is met. According to the method, a tire tread image is photographed by an industrial camera, and then a to-be-processed region of the image is extracted based on grayscale distribution characteristics after being subjected to smooth filtering and threshold segmentation. After that, the pre-preprocessed image is subjected to feature selection, so that the length of the tire tread can be recognized. The lower line segment of the tire tread is extracted by using an edge detection operator, and the included angle thereof relative to a horizontal line segment, namely the tilt angle thereof, can be figured out. The mapping relation between the coordinates of the image and the length thereof is calibrated, and then the actual length of the tire tread can be calculated. The method is fast in detection speed and high in precision.

The embodiment of the invention provides a picture table recognition method and device, electronic equipment and a storage medium, and relates to the technical field of image recognition. The method comprises: carrying out the segmentation of a table region, and obtaining a horizontal line segmentation grey-scale map, a vertical line segmentation grey-scale map, a row region segmentation grey-scale map and a column region segmentation grey-scale map; performing connected region analysis and clustering on the binarized horizontal line segmentation grey-scale map, the binarized vertical line segmentation grey-scale map, the binarized row region segmentation grey-scale map and the binarized column region segmentation grey-scale map to obtain a horizontal line segment object set, a vertical line segment object set, a row region pixel point set and a column region pixel point set; determining a plurality of cells according to the transverse line segment object set, the vertical line segmentobject set, the row region pixel point set and the column region pixel point set; and adding the text content extracted from the picture table into the cells. According to the method and device, theelectronic equipment and the storage medium provided by the embodiment of the invention, the table in the picture table can be accurately identified.

Disclosed is an intersection multi-lane calibration method based on perspective transformation. The intersection multi-lane calibration method comprises the following steps that 1, an image with an obvious lane line is selected from a monitor video of an intersection; 2, edge extraction is performed through a Canny operator on the basis of the selected image, and then n straight line segments are extracted by means of Hough transformation; 3, horizontal line segment sets H and vertical line segment sets S are obtained according to slopes; 4, the sets H and S obtained in the step 3 are screened preliminarily; 5, the H and S finally obtained from the step 4 are marked as H' and S', and the line segments are further filtered; 6, a red light line is determined in H'' and S'' is further filtered so as to determine a lane line; 7, the line segment representing the lane line is found out from the set S''' obtained from the step 6.

A computerized method of recognizing an input hand-drawn table formed by freeform line objects comprises transforming each freeform line object of the table into one of a vertical line segment and a horizontal line segment, generating a grid system based on the vertical and horizontal line segments and converting the generated grid system into a table object.

The invention relates to a phase isolated cast tubular bus and a preparation method thereof and belongs to the technical field of power transmission and distribution. The tubular bus is characterized in that flat joints are arranged at the two ends of a tubular bus main body; a resin insulating layer is cast outside the tubular bus main body; the section outline of the insulating layer is composed of a circular arc segment concentric with the tubular bus, extending line segments tangent with the two sides of the circular arc segment and a horizontal line segment intersecting with the two extending lines; the length of the horizontal line segment is greater than the diameter of the circular arc segment; and the circular arc segments are arranged above to form gas overflowing surfaces during casting and are arranged below to form positioning surfaces during installation. According to the invention, the convention is broken, thus an ideal casting effect can be achieved and the compact insulating casting layer can be formed, without the assistance of special equipment, thus solving the problem of insulating casting of the tubular bus; and the cast planes formed by opening casting become the ideal positioning surfaces during installation, thus achieving two aims at once.

What is disclosed is a system and method for determining whether a front seat in a motor vehicle is occupied based on seat pattern recognition. The present invention takes advantage of the observation that an unoccupied seat of a motor vehicle exhibits features which are distinguishable from an occupied seat. An unoccupied motor vehicle seat typically features long contiguous horizontal line segments and curve segments, and substantially uniform areas encompassed by these segments which are not present in an occupied seat. The present method provides a long horizontal edge test which uses location information within a defined window of the image, edge linking, softness of the edge, number of lines, line / curve fitting, and other techniques to locate horizontal edges in the image which define a seat, and a uniformity step which determines whether the area bounded by the horizontal edges is relatively uniform indicating an unoccupied seat.

The invention belongs to the field of terminals and provides a method and terminal for displaying information. The method includes the steps that two positions, formed by projection on a screen, of two eyeballs of a user are obtained; according to the two positions, a horizontal line segment is determined; a straight line which is perpendicular to the horizontal line segment and passes through the middle point of the horizontal line segment is determined to be a perpendicular line; according to the horizontal line segment, the perpendicular line and the distance between the user and the screen, the information to be browsed is displayed. According to the technical scheme, through changing the displaying position of the information, the possibility of information peeping is lowered, so that user privacy in the terminal is protected.

The invention provides a linear positioning device. The linear positioning device comprises an enclosure as well as a bubble level gauge, an electronic level gauge, and first to third laser parts, and further comprises a computing control module capable of automatically calculating friction, in the total length of a horizontal line segment, of length of a line segment measured by the first laser part or the second laser part, and an information recording module capable of recording fixed fraction, wherein a first laser line and a second laser line are on the same straight line, and are opposite in direction; the first laser line and the second laser line are emitted back to back, and are kept to horizontally emit laser under display guide of the level gauges and under regulating operation of a user; a third laser line is perpendicular to the second laser line and the first laser line, and the third laser line is downwards emitted vertically. The linear positioning device realizes quick and accurate automatic positioning for certain fraction target point on the horizontal line segment, can correspondingly mark out a projection point, on ground, of the certain fraction target point, and further can quickly and accurately locate a target point in a two-dimensional coordinate.

The invention provides a PDF incomplete box line table processing method which comprises the following steps: S1, analyzing a PDF page to obtain elements of the PDF page; s2, judging whether the analyzed elements at least comprise horizontal line segment elements and / or vertical line segment elements or not, and judging whether the PDF page comprises a table or not at least based on the characteristics of the horizontal line segment elements; s3, if the PDF page contains the table, judging whether the table is a complete box line table or an incomplete frame line table at least based on the characteristics of the vertical line segment elements; s4, if the table is an incomplete box line table, obtaining all text blocks in the PDF page and position information of each text block, and obtaining a preliminary table area in the PDF page at least based on the position information of each text block; and S5, based on the horizontal line segment elements and / or the vertical line segment elements, correcting the preliminary table area to obtain a corrected table area. The invention further provides a PDF incomplete box line table processing device, electronic equipment and a storage medium.

A misregistration correcting unit correcting image misregistration on an endless belt in an electrophotographic image forming apparatus is disclosed that includes an image creation part that creates on the endless belt misregistration correction pattern sets each including a pattern of mark groups each being one of a horizontal line mark group formed of horizontal line segment marks and an oblique line mark group formed of oblique line segment marks; a position detecting part that detects the mark positions; a position detection counting part that counts the number of the detected positions of each mark group and detects an abnormal mark group whose number of the detected positions is other than a prescribed number; a detection result storing part that stores information on the detected mark positions; and a misregistration calculating part that calculates the amount of misregistration based on the stored position information excluding that of the abnormal mark group.

A misregistration correcting unit correcting image misregistration on an endless belt in an electrophotographic image forming apparatus is disclosed that includes an image creation part that creates on the endless belt misregistration correction pattern sets each including a pattern of mark groups each being one of a horizontal line mark group formed of horizontal line segment marks and an oblique line mark group formed of oblique line segment marks; a position detecting part that detects the mark positions; a position detection counting part that counts the number of the detected positions of each mark group and detects an abnormal mark group whose number of the detected positions is other than a prescribed number; a detection result storing part that stores information on the detected mark positions; and a misregistration calculating part that calculates the amount of misregistration based on the stored position information excluding that of the abnormal mark group.

The invention relates to a mask plate with an easily removed assistant pattern and a production method thereof. When the assistant pattern is removed, no other auxiliary means is necessary, and the assistant pattern is successfully stripped from a main pattern region with high stripping quality; and the stripped mask plate has a neat and smooth outer frame edge without burrs, and no position deviation is caused during outer frame alignment. The mask plate comprises a main pattern region, an assistant pattern region, and an interval zone; one side of the interval zone is connected with the main pattern region, and the other side is connected with the assistant pattern region; the edge, connected with the main pattern region, of the interval zone comprises at least one non-horizontal line segment retractive relative to the main pattern region, and at least one horizontal line segment; the edge, connected with the main pattern region, of the interval zone is a horizontal line segment; the interval zone is composed of at least one connecting solid bridge, and at least one hollow pattern; the main pattern region is connected with the assistant pattern region through the connecting solid bridge, and parts of the interval zone except the connecting solid bridge are hollow.

The embodiment of the invention discloses a method and device for extracting an effective area of a fisheye image, and the method comprises the steps: determining a horizontal center line and a vertical center line of the fisheye image according to the row number and the column number of the fisheye image; scanning the pixel points on each center line, and determining the boundary point of each center line; and equally dividing a preset number of horizontal line segments and vertical line segments formed by the boundary points, scanning pixel points on each bisector, and determining boundary points of each bisector; selecting any one vertical equally-divided line segment as a bottom edge and any one horizontal boundary point as a vertex, and any one horizontal equally-divided line segment as a bottom edge and any one vertical boundary point as a vertex to construct a triangle; and extracting an effective region from the fisheye image according to the coordinates of the center of the circumscribed circle of each triangle and the radius value. The preset number is obtained according to the fact that the effective area calculated amount is smaller than the preset value and the effective area error is smaller than the preset error value, and the effective area extraction efficiency is improved while the accuracy is guaranteed.

The invention provides a document table analysis method and device, and relates to the technical field of file analysis. The method comprises the steps of obtaining line segment feature information related to a table to be identified in a non-edited document, wherein the line segment feature information comprises coordinate information of line segments; obtaining horizontal line segment information and vertical line segment information in the to-be-identified table according to the line segment feature information; obtaining initial form information according to the horizontal line segment information and the vertical line segment information in the form to be identified; and generating an editable table corresponding to the table to be identified according to the initial table informationand the coordinate information of the line segments. By obtaining horizontal line segment information and vertical line segment information in a table to be identified through line segment feature information, obtaining initial table information according to the horizontal line segment information and the vertical line segment information, an editable table corresponding to the table to be identified is generated according to the initial table information and the line segment feature information related to the table to be identified. According to the method, the problems of high restoration complexity and poor restoration effect of the composite table and the missing table are effectively solved.

The invention discloses a steel structure beam-column joint haunching reinforcing structure and a manufacturing method thereof. A steel structure beam-column joint is composed of a steel column and a steel beam and is reinforced through a haunch. The haunch is formed by a haunch flange and a haunch web in a welding connection and reinforcing mode. The two ends of the haunch flange are provided with welding grooves, and the haunch web is provided with a first overwelding hole, a second overwelding hole and a clearance hole. The boundary, on the haunch web, of the first overwelding hole is composed of a horizontal line segment, an arc and an oblique line segment. The boundary, on the haunch web, of the second overwelding hole is composed of an arc. The boundary, on the haunch web, of the clearance hole is composed of a horizontal line segment, an arc and an oblique line segment. The haunch flange and a column flange are connected through a first welding line, the haunch flange and a beam flange are connected through a second welding line, the haunch web and the column flange are connected through a third welding line, and the haunch web and a beam web are connected through a fourth welding line.

An aspect of the present invention provides an image processing apparatus including: a people detecting unit that detects image portions of faces of a plurality of people in an image; and an image converting unit that enlarges a portion below a horizontal line segment of the image in a manner that locates the horizontal line segment at a top end of the enlarged image based on upper end positions of the detected faces. The horizontal line segment is on or above the extracted faces and has a lateral width equal to or larger than a lateral width of the faces. The image processing apparatus performs enlargement with reference to the faces so as not to warp the faces.

The invention discloses a wide-speed-range aircraft waverider configuration design method, which comprises the following steps: determining a free incoming flow condition based on two-stage injection stage separation, a reference cone length and a shock wave angle for designing a waverider, and obtaining a supersonic-speed circular cone reference flow field model required for designing the waverider; according to two-stage injection stage separation requirements, initial parameters of half waverider are determined, and the whole waverider is obtained through mirror symmetry of the half waverider; according to the two-stage injection stage separation requirement, the datum line type of the half waverider is determined, the datum line type of the half waverider specifically comprises a horizontal line segment, a lower opening parabolic segment and an upper opening parabolic segment, and datum line type functions of the horizontal line segment, the lower opening parabolic segment and the upper opening parabolic segment are obtained according to the initial parameters; according to the initial parameters and the datum line linear function, the wave rider base body is obtained through calculation based on the supersonic speed surrounding cone datum flow field model needed for designing the wave rider by means of the cone guided shock wave theory, and the configuration process is simple and efficient.

The invention discloses a depth information measurement method based on a single-frame composite network template, and mainly solves the problems that an existing time coding method consumes too longtime, cannot dynamically measure scene depth information and is low in depth information density. According to the scheme, the method comprises steps of designing a single-frame composite network template P; projecting P to a target object, and shooting the target object to obtain a deformed image; transmitting the deformed image to a computer so as to separate a codeword graph I '2 and a line segment grid graph I' 1 from the deformed image; performing line extraction by utilizing the separated line segment grid chart I '1 to obtain a line segment grid chart; decoding according to the separated codeword graph I '2, and solving a line segment matched with the template by decoding the codeword at the intersection point of the horizontal line segment and the vertical line segment; and calculating the depth information value of the to-be-measured object according to the line-plane intersection principle by utilizing the corresponding relation between the line segments. The method can dynamically obtain scene depth information, is high in data precision, and can be used for industrial monitoring, medical science, man-machine interaction, virtual reality and 3D printing.