41 results about "Energy function minimization" patented technology

In a novel method of 3D modeling of an object from a video sequence using an SfM algorithm and a generic object model, the generic model is incorporated after the SfM algorithm generates a 3D estimate of the object model purely and directly from the input video sequence. An optimization framework provides for comparison of the local trends of the 3D estimate and the generic model so that the errors in the 3D estimate are corrected. The 3D estimate is obtained by fusing intermediate 3D reconstructions of pairs of frames of the video sequence after computing the uncertainty of the two frame solutions. The quality of the fusion algorithm is tracked using a rate-distortion function. In order to combine the generic model with the 3D estimate, an energy function minimization procedure is applied to the 3D estimate. The optimization is performed using a Metropolis-Hasting sampling strategy.

The invention relates to a KNN (K-Nearest Neighbor) sorting algorithm based method for correcting and segmenting the grayscale nonuniformity of an MR (Magnetic Resonance) image, belonging to the field of image processing. The method comprises the following steps of: firstly constructing a grayscale nonuniform field model by utilizing surface fitting knowledge and using a group of orthonormalization basis functions, and establishing energy functions; and then solving model parameters according to an energy function minimization principle to realize grayscale nonuniformity correction and image segmentation, wherein subordinate functions are solved by adopting an iterative algorithm and the KNN algorithm in the model parameter solving process, therefore a partial volume effect is greatly reduced while a grayscale nonuniform field is eliminated, and the influence of noises on the correction and the segmentation of the grayscale nonuniformity of the MR image is reduced. The subordinate functions are solved with KNN through the following steps of: firstly acquiring an accurate smooth normalization histogram by using a kernel estimation algorithm; then respectively solving a threshold value TCG between cerebrospinal fluids and gray matters and a threshold value TGW between the gray matters and white matters by using a maximum between-cluster variance method; carrying out rough sorting on the KNN sorting algorithm by utilizing the two threshold values; and finally accurately sorting points to be fixed by adopting the traditional KNN sorting algorithm.

A method registers a source image with a target image, wherein the images are deformable, by first measuring dissimilarity between the source image and the target image. The dissimilarity minimized using a discrete energy function. At multiple scales, multi-scale Markov random field registration is applied to the source and target images to determine a deformation vector field. Then, the target image is warped according to the deformation field vector to obtain a warped target mage registered to the source image.

Methods, systems, and computer-readable media for reconstruction a three-dimensional scene from a collection of two-dimensional images are provided. A computerized reconstruction system executes computer vision algorithms on the collection of two-dimensional images to identify candidate planes that are used to model visual characteristics of the environment depicted in the two-dimensional images. The computer vision algorithms may minimize an energy function that represents the relationships and similarities among features of the two-dimensional images to assign pixels of the two dimensional images to planes in the three dimensional scene. The three-dimensional scene is navigable and depicts viewpoint transitions between multiple two-dimensional images.

The invention discloses an image vehicle attitude estimation method based on a fine three-dimensional model. The method comprises steps: first, the position of the vehicle being detected from the image, and initializing the vehicle three-dimensional posture parameters, rendering the vehicle three-dimensional model to the image plane using the current posture parameters, extracting the contours ofthe image vehicle and the model vehicle respectively, using the matching error between the contours to construct the energy function, and adopting the Gauss-Newton algorithm to optimize the posture parameters, solving the energy function minimization problem, and obtaining the final result. In contour matching, the improved random sampling consistency (RANSAC) algorithm is used to fit the contourof the curve model with piecewise straight lines, and then the contour is matched with the real vehicle contour. The invention can accurately and robustly recover the three-dimensional posture parameters of the vehicle relative to the camera in the surveillance video in the complex surveillance scene, which is of great significance to the understanding and automatic processing of the surveillancevideo.

The invention relates to the technical field of computer vision. The embodiment specifically discloses a method and a system for three-dimensional reconstruction of a human face based on binocular stereo and photometric stereo. The initial depth image of a target human face is obtained by binocular stereo matching. Four images of the target face collected by two light sources and two camera devices are represented by photometric stereo as constraint terms and combined with the initial depth image to construct the energy function of the depth image to be solved. The optimized depth image is obtained by minimizing the energy function, and the optimized depth image is transformed into a three-dimensional point cloud to obtain the three-dimensional point cloud of the target face. It solves theproblem that the existing photometric stereo vision needs at least three light sources for 3D reconstruction, and the scale-free feature has deformation, and realizes the accurate 3D reconstruction of human face without deformation.

The invention belongs to the technical field of image processing, and provides a binocular stereo matching method and apparatus based on oblique plane smoothing. The method includes: performing subsequent optimization on a semi-global stereo matching algorithm by employing an oblique plane smoothing algorithm; and segmenting an image by employing the thought of energy function minimization, classifying boundary points, and performing oblique plane smoothing on parallax fracture boundary points and shielding boundary points according to conditions. According to the binocular stereo matching method based on oblique plane smoothing, the phenomenon of parallax fracture can be solved, the final effect of a disparity map is smooth, and good smoothing effect can be achieved in shielding regions, low-texture regions, and depth discontinuous regions.

The invention provides a robustness prospect detection method based on multi-view learning. The method includes the steps that a reference background image is acquired by an input video through time domain median filtering method, and iterative search and multi-scale fusion are conducted on a current image and the reference background image to acquire heterogeneous characteristics; conditional probability density of prospects and conditional probability density of backgrounds are calculated by using condition independence of the heterogeneous characteristics, and a prospect posteriori probability and a background posteriori probability are calculated by using Bayes rules according to prospect likelihood, background likelihood and a prior probability; an energy function of a Markov random field model is established by means of the prospect posteriori probability, the background posteriori probability and space-time consistency constraint, the energy function is minimized by using a belief propagation algorithm, and segmented results of the prospect and the background are obtained. By means of the method, in a complex challenging environment, robustness prospect detection is achieved.

The invention relates to a human upper body detection and splitting method applied to a low-contrast video. The method mainly comprises two processes. In the first process, a communicated area representing a foreground object is extracted from a current frame by a background subtraction technology and a morphological method; and for each foreground area, the shape features of a polar-coordinate-based two-dimensional histogram corresponding to the foreground area are extracted as the input of a pre-trained support-vector-machine-based classifier, and a class tag corresponding to a human upper body class and a class tag corresponding to a non-human upper body class are output. In the second process, when an area which is identified as a human body area is misjudged as a non-human body area,the area is represented by an energy function, an inaccurate contour line is corrected by an energy function minimization process at the same time, and finally, a background frame is updated on the basis that an accurate foreground human body contour is obtained. By the method, a video with low contrast and resolution can be processed in real time, and both detection accuracy and a splitting result can meet the requirements of application.

The invention discloses a fan blade image segmentation and grid optimization splicing method. The fan blade image segmentation and grid optimization splicing method comprises the following steps: S1,continuously collecting images of a single-sided fan blade to form a group of to-be-spliced original images; S2, using the U-net algorithm to carry out image foreground segmentation on each original image, and extracting a fan blade main body part to form a group of images to be registered; S3, gridding each image to be registered, establishing an index from 1 to m for a grid vertex of each to-be-registered image, then expressing x and y coordinates of the grid vertexes as a vector V of a 2m dimension, defining a global energy function about V, and minimizing the energy function to obtain a grid vertex optimal solution, and S4, completing image splicing according to the grid vertex optimal solution. The method has the advantages that the problem that target splicing fails due to the fact that the depth of field of the target and the background is too large is solved, natural splicing of multiple fan blades is achieved, and spliced images are small in visual effect distortion, continuous and real.

The invention comprises: a image memory 1, a component for executing the high level vision 2, a basic polygon transform component 3, an orthogonal location transform component 4, a free difference operation component 5, a target outline extracting component 6, and an energy function minimizing component 7.

The invention discloses a ceramic prototype three-dimensional mesh model generation method and device based on multi-view sketching, wherein the method comprises the following steps: acquiring ceramicprototype sketches of a plurality of orthogonal views, sequentially sampling the ceramic prototype sketches of the plurality of orthogonal views one by one, and acquiring the segment-by-segment sampling results; constructing an energy function according to the segment-by-segment sampling results, and obtaining the modeling curve by a minimization method; conducting three-dimensional mesh fillingon the modeling curve, and outputting the result. This method takes advantage of the characteristics of ceramic prototype design, takes the sketch of multiple views as input, and constructs the three-dimensional mesh model by correlation sampling and energy function minimization. The method improves the efficiency of ceramic prototype construction and is simple and easy to implement.

The invention discloses a method for reconstructing an object surface and a device thereof and a computer storage medium, and the method comprises the steps: generating a corresponding dense point cloud through a plurality of images, and the plurality of images can be obtained through photographing a scene from different photographing position points; based on the dense point cloud, generating a tetrahedral mesh corresponding to the dense point cloud, the vertex of each tetrahedron in the tetrahedral mesh being a coordinate point in the dense point cloud; generating a binary label for each tetrahedron based on energy function minimization, the binary labels being used for representing that the tetrahedrons are located inside or outside the surface of the object; and extracting a common surface between the tetrahedrons with different binary labels to reconstruct the surface of the object. The energy function comprises the sum of first penalty terms corresponding to the tetrahedron and the sum of second penalty terms corresponding to the common surface, and the second penalty terms comprise grid density weights. Through the above mode, the negative influence of the noise points on the surface details of the reconstructed object is reduced, and the reconstruction precision is improved.

Disclosed in the invention is a total variation and euler elastic rod-based supervised mode identification method, comprising: constructing a total variation and Euler elastic rod-based energy function under the framework of least square regularization; utilizing a variational method to convert solution of energy function minimization into solution of a corresponding Euler-Lagrange differential equation; solving the differential equation to further obtain a final classifier; and carrying out mode identification on data by utilizing the classifier. According to the invention, a novel method is provided for solution of a supervised mode identification problem and in general, can be applied to solution of a classification problem like handwriting digit identification; and the provided method enables an effect that is comparable with one caused by an existing popular method to be realized for most of data sets.

The invention relates to a method for obtaining a geometric mass center of an object, and the method is used for the object formed by a 3D manifold. The method comprises that the internal distance of the 3D manifold is calculated; an energy function model aimed at the 3D manifold is constructed, an energy function is minimized by using a gradient descent method, and a gradient function corresponding to the 3D manifold is obtained; according to the obtained gradient function of the energy function of the 3D manifold, a shortest path between two points in the 3D manifold is obtained; and N different initial values are selected randomly from points in the 3D manifold, the energy function is optimized, an optimal energy function value is obtained, and a point corresponding to the optimal energy function value serves as the geometric mass center of the object. The method can ensure that the mass center is positioned inside the 3D manifold and is not influenced by the appearance of the object, and adverse influence of noise and deformation can be avoided.

The invention provides a multiple-motor coordinated control method based on electric locomotive traction total quantity uniformity. The method includes that a multiple-motor coordinated control framework uniform in total quantity is built, a multiple-traction system distribution strategy based on energy function minimization is provided, and multiple motors are enabled to complete output torque adjustment by means of optimal dynamic adjusting. A sliding mode variable structure controller is established aiming at each independent traction subsystem, and actual output torque of each motor is tracked and fed back. The control method is high in interference ability, good in tracking performance and capable of still maintaining output total equantity and uniformity in total quantity needed fortotal quantity, and basically constant traction total amount in the process of electric locomotive running is ensured.

The invention discloses a non-linear control method for an on-board charger PFC converter based on discrete energy function control, which includes the steps of sampling input voltage and current, detecting peak input power, and synthesizing a load reference value and an input power reference value according to the peak input power; building a discrete state space model of an interleaved parallel PFC Boost converter according to the Euler discretization method; and designing a discrete energy function controller of the interleaved parallel PFC Boost converter system according to the discrete energy function minimization control method. The invention proposes an algorithm based on the discrete energy function control, which effectively solves the problem of slow dynamic response of PFC under the original control algorithm. In addition, the non-linear control method simplifies the design of the converter, improves the control effect of the system and has good flexibility and steady-state output characteristics.