41 results about "Vector flow" patented technology

A system and method for segmenting cardiac images and, in particular, segmenting the left ventricle of the heart using a contour propagation model that integrates visual information and anatomical constraints. The visual information comprises a gradient vector flow-based boundary component and a region component that separates the cardiac contours / regions according to their global intensity properties that reflect the different tissue properties. The anatomical constraints couple the propagation of cardiac contours according to their relative distance. The propagation model comprises a weighted integration of a boundary segmentation model, a region model and coupling function.

The invention discloses a liver multi-phase CT image fusion method which comprises the steps of firstly performing coarse registering on a source image sequence according to a multi-resolution CT image registering method based on a combined histogram, then realizing automatic liver image dividing according to a region growing algorithm in combination with confidence connection and liver image dividing based on a gradient vector flow snake model, and effectively extracting the edge information of the liver; performing blood vessel extraction based on a directional region growing algorithm on the liver image, performing free deformation transformation based on a B spline and liver non-rigid registering based on space weighting mutual information on a liver essential image, thereby accurately finding an image pair at a same position of a space; and finally performing image fusion based on wavelet transformation. The liver multi-phase CT image fusion method aims at the characteristics of a liver CT image. An image separation process and an image registering process are combined with the image fusion process, thereby greatly improving fusion precision.

The utility model discloses an extracting method of building contour in a high resolution remote sensing image of satellite, comprising the following steps: (1) the image is pretreated to improve the definition of the image; (2) a parameter of the split minimum area is defined by the size of the extracted building, the split of the high solution remote sensing image is performed based on the objective; (3) the edge of the split image is fitted using the dynamic contour model calculation based on the gradual vector flow; (4) vectorization; (5) optimizing processing. The utility model has advantages that the method is able to reflect the actual contour of the building and the output is of vector format, which is able to update the vector data of GIS directly.

The invention relates to a positioning method and device for tables in a PDF document. The method comprises the steps that the PDF document containing the tables is received; character information andstraight line information are extracted from vector flow information of the PDF document; and according to the extracted character information and straight line information, a table region in the PDFdocument is positioned. According to the positioning method and device, table region positioning is performed based on all straight lines and text blocks in pages; and compared with the prior art, the accuracy of table region positioning can be improved, and consequently a basis is provided for accurate analysis of table information.

Embodiments of the invention utilize a variational framework for computing curve skeletons (CS) of objects whose cross section is not necessary tubular. Embodiments utilize an energy function, which is proportional to some medialness function, such that the minimum cost path between any two medial voxels in the shape is a curve skeleton. Different medialness functions include the Euclidean distance field and a modified version of the magnitude of the gradient vector flow (GVF), which results in two different energy functions. The first energy function controls the identification of the topological nodes of the shape from which curve skeletons start, while the second one controls the extraction of the curve skeletons. Preferred embodiments are completely automated since all parameters are analytically estimated. Embodiments are highly less sensitive to boundary noise, are able to extract the entire curve skeletons, as well as only part of it given the starting and the end voxels, and do not require voxels to be isotropic. In addition, computed curve skeletons are highly centered and form a connected graph. Preferred embodiments have been validated the framework both quantitatively and qualitatively against several 3D shapes of different complexity.

The invention provides a target tracing method which is based on a particle filter and active contour model, and relates to a target tracing method for a gradient vector flow-parameter active contour model. The method firstly adopts a background difference method to obtain target initial profile, and the improved gradient vector flow-parameter active contour model is utilized, thus leading the parameter active contour model to be converged to true profile of a moving target; control points are increased or reduced according to the distance of the control points, so as to achieve the purpose of tracing moving and morph targets in a self-adapting way; then energetic particle filter target tracing algorithm is used for tracing the targets by combining the particle filter and the improved gradient vector flow-parameter active contour model; and tracing strategy when the target is sheltered is used, thus overcoming the effect of sheltering in the tracing process. In a whole tracing process, profile point information is not sheltered mostly by fully utilizing GVF-Snake model target profile, so as to effectively overcoming the effect such as complex environment and the like in the tracing process.

The invention relates to a multiple target tracing method which is based on a particle filter and active contour model, in a whole tracing process, profile point information is not sheltered mostly by fully utilizing GVF-Snake model target profile, so as to effectively overcoming the effect such as complex environment and the like in the tracing process; after obtaining the target initial profile, an improved gradient vector flow-parameter active contour model is utilized, thus leading the parameter active contour model to be converged to true profile of a moving target, and control points are increased or reduced in a self-adapting way according to the distance of the control points; the multiple target is traced by combining the particle filter and by using improved K typical value clustering algorithm and energetic particle filter; tracing strategy when the target is sheltered is used, thus overcoming the effect of sheltering in the tracing process.

A propulsion linearizing mechanism is provided for linearizing a fluid flow. The mechanism includes a frame having a cylindrical outer baffle which rotatably supports a plurality of propeller elements thereon. Each propeller element defines a respective sweep area as the propeller element is rotated which overlaps sweep areas of adjacent propeller elements. The outer baffle circumscribes an outer periphery of the collective sweep areas of the respective propeller elements. The propeller elements rotate in the same direction whereby forces of curvature flow of adjacent propeller elements substantially cancel one another to linearize fluid flow through the mechanism. Additional baffles and infills within the areas of non blade sweep may be provided for particular applications of the mechanism. In various applications, linear forces of vector flow are formed by integrating curvature forces of tangential flow and economy flow systems are established to increase force potentials on the planes of rotating propellers to provide the emission and induction flow with an insulation whereby fluid in the immediate vicinity of the mobile flow remains in an undisturbed static state. This allows a fluid propulsion assembly to be fitted with an outer utility mantle in the static zone of the field.

The invention discloses a method for tracking a video target. The method comprises the following steps that: gradient vector flow (GVF) deformation is carried out to each candidate position in a current frame to obtain each deformation curve; the video characteristic of the deformation curve is calculated; according to the calculated video characteristic, a candidate position is determined as a target position. The invention also discloses a video target tracking device and an automatic video tracking system, which can improve the accuracy of target tracking.

The invention discloses a human body thoracic and abdominal cavity CT image aorta segmentation method based on a GVF Snake model. The method overcomes the shortcomings of the heavy workload and long time consuming of the traditional manual and semi-automatic segmentation, the repeatability of the method is good, and the uncertainty caused by artificial segmentation is prevented. The method includes (1) reading a CT image and performing image preprocessing; (2) performing the initial profile setting of the GVF Snake model on the image obtained after the preprocessing; (3) obtaining the edge image of the image after the preprocessing; (4) obtaining gradient vector flow GVF as the external energy field by the diffusion equation based on the obtained edge image; (5) establishing an internal energy model to maintain the smoothness of the profile; and (6) constructing an energy function E by means of internal energy and external energy, obtaining the minimum value of energy E by means of iteration operation, and the target boundary of the profile can be obtained at the end. The method has important application values in the field of human body thoracic and abdominal cavity aorta interlayer segmentation diagnosis treatment.

The invention discloses a medical image segmentation method based on an improved gradient vector flow model. The medical image segmentation method comprises the following steps: carrying out pretreatments on a fuzzy medical dermoscopy image, wherein the pretreatments include image equalization, wave filtration, edge detection and the like; firstly extracting a rough profile of the image based on a texture segmentation method, and acquiring valid information points of the image; replacing the calculation of a European space vector field with a method for calculating a gradient vector field in a Riemann space, thereby improving the calculation efficiency; and finally, carrying out fine treatment on the image by virtue of an improved gradient vector snake model, and extracting an accurate and smooth image profile, thereby providing reference to medical diagnosis. According to the medical image segmentation method, the rough profile of the dermoscopy image is extracted by virtue of the texture segmentation method, and the gradient vector flow model calculated by a gradient force field is improved, so that the profile of the fuzzy medical dermoscopy image can be accurately and rapidly extracted, and a segmentation result is displayed and provides a contrastive reference for the medical diagnosis.

The invention discloses a multi-operator image redirection method for maintaining the important content area width-to-height ratio. Firstly visual saliency detection is performed on an input image, and an energy spectrum is calculated by combination of a normalized grayscale map; a visual saliency image is obtained by performing threshold segmentation and division on the energy spectrum; the required stretching size is calculated according to the size of the visual saliency image for reducing the overall size of the visual saliency image; different degrees of stretching is performed on the visual saliency image and background information; then similarity transformation is performed on the stretched image; and finally fine adjustment is performed by using a line cutting method with addition of gradient vector flow according to the target size so as to acquire the final output image. With application of the method, the visual saliency content can be detected and globally protected so as to enhance the quality of the image redirection technology.

The invention relates to a method for segmenting a cardiac nuclear magnetic resonance image. The method for segmenting the cardiac nuclear magnetic resonance image comprises the following steps of: 1, performing Gaussian filtering preprocessing on the acquired cardiac nuclear magnetic resonance image; 2, computing an external force field of a generalized gradient vector flow based on expansion neighborhood and noise smoothing on the preprocessed image; 3, defining an initialized outline position of the left ventricular endocardium of the heart; 4, segmenting the left ventricular endocardium of the heart; 5, defining a final segmenting outline result of the left ventricular endocardium of the heart into an initialized outline position of the left ventricular epicardium of the heart; 6, setting the boundary strength of an area surrounded by the endocardium outline in an original boundary graph to be 0, recomputing the external force field of the generalized gradient vector flow based onthe expansion neighborhood and the noise smoothing; and 7, segmenting the left ventricular epicardium of the heart. Based on convolution computation, and by taking energy constraint of an elliptical shape, the method for segmenting the cardiac nuclear magnetic resonance image has the advantages of high computing speed, wide capturing range, strong anti-noise ability, excellent performance in weakboundary protection and deep dented region segmentation, and capability of accurately segmenting the left ventricular endocardium and the left ventricular epicardium of the heart.

The utility model discloses a method to detect the lake shore line and extract the lake contour from the synthetic aperture radar remote sensing image. The utility model adopts the technical proposal that: firstly, an all-direction self-adaptive dynamic window filter is utilized to do speckle noise suppression on the synthetic aperture radar remote sensing image; secondly, the fact that the singularity of the edge and the noise have different propagation laws according to scale variation in dyadic wavelets transformation is utilized to detect edge points; thirdly, the traditional contour following method is utilized to connect the edge points; finally, an active contour model based on gradient vector flow is utilized to fit the lake contour. The utility model effectively eliminates the influence of speckle noise in SAR image on flood boundary extraction, detects the lake shore line accurately and finally gets the lake contour.

An object detection method with a rising classifier effect is embedded in an object detection device and has steps of acquiring image information; determining a position of an obstacle in the image information, wherein the image information is detected for generation of an extracted range corresponding to the obstacle; recognizing an extracted contour of the obstacle, wherein the extracted contour of the obstacle associated with the extracted range is obtained by using an algorithm for Poisson gradient vector flow based active deformable contour model and a multi-mesh algorithm with different grid sizes; and forming a maximum border of the extracted contour of the obstacle for a classifier to recognize a type of the obstacle according to the maximum border of the obstacle. Accordingly, the object detection method is advantageous in complete extraction, higher recognition rate, faster computation and feasibility to be integrated with vehicle systems.

The invention relates to a cardiac nuclear magnetic resonance image segmentation method which comprises the following steps of: 1, carrying out Gaussian filtering pretreatment on an image; 2, calculating an external force field of an edge for keeping a normal gradient vector flow; 3, defining an initialization outline position of an inner membrane of the cardiac ventriculus sinister; 4, adding a circular energy constraint in a curve evolution process, and segmenting to obtain the inner membrane; 5, defining a final segmentation outline result of the inner membrane as an initialization outlineposition of an outer membrane; 6, setting the edge strength of a zone enclosed by the inner membrane outline in an original edge graph as 0, and re-calculating the external force field; and 7, addinga circular energy constraint in a curve evolution process, and segmenting to obtain the outer membrane. The invention has the advantages of large capturing range, strong noise proof capacity, better robustness to weak edge leakage, and capability of accurately segmenting the inner membrane and the outer membrane of the ventriculus sinister wall.

The application provides a method and a device for extracting a vascular ridge point based on an image gradient vector flow field. The method comprises steps that the blood vessel target in the angiographic image is enhanced; the gradient vector flow field model is used to obtain the gradient vector flow field of the enhanced image; the cosine of the vector at a pixel in the enhanced image and the vector angle of the two adjacent pixels in the direction of the coordinate axis is calculated according to the gradient vector flow field of the enhanced image, if the cosine satisfies the threshold condition, the pixel point is the vascular ridge point in the direction of the coordinate axis. The present invention can effectively suppress the influence of background noise, improve the extraction efficiency of vascular ridge points in angiographic images, increase the detection number of vascular ridge points and improve the detection precision of vascular ridge points, and provide the basis for the subsequent blood vessel centerline extraction and blood vessel modeling.

The invention is applicable to the field of computer technologies, and provides a method and device for extracting a center line of a tubular target. The method comprises the steps of preprocessing an image of a tubular target center line to be extracted to obtain a corresponding enhanced image, acquiring a gradient vector flow field of the enhanced image, acquiring deformation force parameter corresponding to an initial ridge line segment list according to the gradient vector flow field, extracting tubular target ridge points from the enhanced image, generating the initial ridge line segment list, building a regularized open curve deformation model, performing deformation and processing on a ridge line segment in the initial ridge line segment list according to the model before a deformation ending condition is met to obtain a corresponding tubular target center line segment, updating the initial ridge line segment list according to ridge line segments traversed in the deformation process, and generating a center line of the tubular target according to the tubular target center line segments when the deformation ending condition is met, thereby optimizing the deformation process of the ridge line segments, reducing the deformation time of the ridge line segments and improving the extraction accuracy and the extraction rate of the center line.

The invention provides an X-band radar ocean current inversion method based on cross-spectrum analysis, which comprises the steps of firstly respectively performing cross-spectrum analysis on two adjacent images in an X-band radar image to obtain a coherent coefficient spectrum and a phase spectrum; then respectively averaging the coherent coefficient spectrum and the phase spectrum, determining the number of dominant waves according to the peak value of the coherent coefficient spectrum, and eliminating the direction ambiguity of the dominant wave direction by using the average phase; selecting different wave directions, establishing a mode according to the phase velocity obtained from a frequency dispersion relation, and solving the model according to a least squares method to obtain anocean current vector. The X-band radar ocean current inversion method can solve the problems of low accuracy of the X-band radar in observing sea surface flow field and inability of obtaining the vector flow field by using a set of coherent X-band radar in the prior art, thereby meeting the requirements of business-oriented observation for the coastal marine environment and services for marine production and life and the like.

The invention discloses an fingerprint image direction field calculation method in an automatic fingerprint identification system. The fingerprint image direction field calculation method adopts a GVF-model based method and comprises the following steps: (1) inputting a fingerprint image; (2) calculating the gradient distribution, the quality score and the crack mask of the input fingerprint image; (3) judging the fingerprint quality, carrying out the step (4) if the fingerprint quality is bad, or carrying out the step (5) if the fingerprint quality is not bad; (4) actuating the crack mask and the initial GVF (gradient vector flow) field V0, and omitting the noise data in a crack region; (5) diffusing the initial GVF field V0 rapidly and smoothly, and taking the diffused GVF field as the Vcoarse; (6) calculating the distribution theta corse (x, y) of the direction field of the fingerprint by using the Vcoarse, detecting a singular point, and obtaining a singular region mask MSingulsr; (7) mending the Vcoarse of a singular region by using a GGVF (general gradient vector flow) model to obtain a final GVF field VFine; and (8) calculating the final direction field theta (x, y). The method overcomes the detects of poor low-quality background region image calculation effect of the prior art and achieves good low-quality fingerprint image calculation effect and the good result of obtaining the direction filed in the fingerprint background region.

The present invention discloses a medical sequence image motion estimation method based on the generalized fuzzy gradient vector flolw field, it includes the following steps: 1. obtaining and sequence image; 2. obtaining generalized fuzzy gradient vector flow field of double-step tracking model, and locak correlation of generalized fuzzy gradient vector flow field and optical flowvector field; 3. manually drawing the outline of edge of interested zone of first frame image; 4. under the action of three kinds of external force fields using said double-step tracking model to frame-by-frame track the drawn edge outline of interested zone; 5. combining the above-mentioned tracking result of outline, and using maximum posteriori estimation to make optimizational estimation and tracking of every point on the outline so as to obtain the optimum motion track of the point.

The invention discloses a method for segmenting an aortic-valve ultrasound image sequence based on an interframe-shape-constraint GCV model. The method includes the following steps: acquiring a group of ultrasound images and carrying out Wiener filtering; calculating a gradient vector flow field of each image frame by frame and adding the gradient vector flow fields to a CV model as energy constraints and obtaining the GCV model; through defining an initial constraint shape and adding the initial constraint shape into the GCV model as an energy constraint item and then minimizing an energy functional so that a segmentation result of a first frame of image is obtained; and carrying out filtering of a rolling sphere method on the aortic-valve segmentation result of an adjacent previous-frame image and adding the obtained result into the GCV model as an energy constraint item so that the segmentation result of a current frame is obtained through calculation. The method for segmenting the aortic-valve ultrasound image sequence based on the interframe-shape-constraint GCV model is operated on short-axis images of ultrasonic cardiograms so that not only is work load of doctors reduced, a problem that serious overflow in segmentation of the aortic-valve ultrasound images in the prior art is also solved. The segmentation result is extremely close to the result of manual segmentation. The method enables the aortic valve to be segmented in a simple and highly-efficient manner.

The invention discloses a GVF Snakes (gradient vector flow snakes) model based iris location algorithm. The GVF Snakes model based iris location algorithm includes 1), subjecting an original iris image to coarse positioning and looking for a pseudo circle of a pupil; 2), calculating the GVF value of the image and subjecting the inner edge of an iris to coarse positioning by the aid of a GVF Snakes module; 3), subjecting the inner edge of the iris to accurate positioning by the aid of the step 2) to determine the geometrical parameter of the inner edge; 4), detecting the outer edge of the iris by the aid of the GVF Snakes module, processing the outer edge of the iris as a circle and positioning the outer edge of the iris; 5), determining the geometrical parameter of the outer edge of the iris by the aid of the outer edge of the iris positioned in the step 4); 6), drawing a circle by the aid of the center and the radius of the outer edge, wherein the circumference is an accurately positioned image of the outer edge of the iris, and the iris can be accurately positioned by combining the center and the radius of the inner edge. The GVF Snakes model based iris location algorithm is a new method in positioning of the iris and has the advantages of high adaptive capability and accurate algorithm, the GVF Snakes model is easy to implement with new constraints easily added and the like.

The invention discloses a method for segmenting textile and medicine images, which improves the over-segmentation in an image watershed algorithm and comprises the following steps of: firstly, carrying out boundary detection on an original image for obtaining a gradient amplitude image; secondly, carrying out boundary information diffusion and noise removal on the gradient amplitude image by using a one-dimensional gradient vector flow partial differential equation; thirdly, detecting local minimum value points of the gradient image, and automatically merging similar local minimum value points through morphology dilatation operation; and finally, segmenting the processed image by using the watershed algorithm.

The invention relates to a medical image segmentation method, in particular to a medical image segmentation method used for a liver model in a robot virtual training system. The medical image segmentation method used for the liver model in the robot virtual training system is disclosed to solve the problem that no corresponding virtual liver models established according to patient liver features exists at present. The medical image segmentation method is implemented by the steps of: step 1, acquiring an effect image after filtering; step 2, selecting a kernel function image segmentation method for pre-segmenting the effect image after filtering to obtain a pre-segmented image; step 3, and utilizing an improved semi-automatic gradient vector flow GVF_Snake method to realize segmentation of the pre-segmented image. The medical image segmentation method is applied to the field of medical image segmentation of the liver model in the robot virtual training system.

An ultrasound imaging system (100) includes a velocity processor (118) that processes ultrasound data representing structure flowing through a tubular object and generates vector flow imaging information indicative of the structure flowing through a tubular object based thereon. The vector flow imaging information includes an axial velocity component signal and a lateral component signal, and the axial and lateral component signals indicate a direction and a speed of the structure flowing through the tubular object. The ultrasound imaging system further includes a flow parameter processor (120) that determines at least one flow parameter based on the vector flow imaging information and generates a signal indicative thereof.

The invention discloses an image segmentation method based on appearance dictionary learning and shape sparse representation. The method comprises the following steps: moving an average grid to an initial central position; positioning the position of each mark point on the grid along the normal direction of the average grid; obtaining an initial segmentation result in combination with a shape sparse representation algorithm; using an algorithm combining appearance dictionary learning and normal searching to adjust mark points nearby the obtained initial segmentation result; performing constraint reconstruction on the adjusted grid by using a shape sparse representation algorithm again; adjusting the mark point again according to the reconstruction result in combination with the characteristics of the gradient vector flow field and the probability value of label reconstruction of appearance dictionary learning; and a final segmentation result is obtained by using a shape sparse representation algorithm. According to the method, the appearance prior information with the resolution capability and the shape prior information with the reconstruction capability of shape sparse representation are learned by fully utilizing the appearance dictionary, so that the mark point positioning algorithm and the sparse shape representation algorithm are complementary, and finally a more accuratelow-dose CT segmentation result is obtained.