585 results about "Transformation algorithm" patented technology

A method, medium, and apparatus extracting avatar parameters using an avatar transformation algorithm from health state information measured from a user and generating an avatar using the extracted parameters. In the apparatus, a state analysis unit receives at least one among a bio-signal and questionary information of a user and outputs health state information, a parameter extraction unit extracts parameters defining an avatar image, and an avatar generation unit generates the avatar image using the parameters. By providing the method, medium, and apparatus, it is possible to provide a user with a convenience in recognizing his health state by generating an avatar image describing the user's future health state as well as the user's current health state. Further, it is possible to induce a user to manage his health constantly by maintaining a user's with the avatar image change.

The design, internal data representation and query model of the invention, a hierarchical distributed caching system for semi-structured documents based on LDAP technology is presented that brings both, the semi-structured data model and the LDAP data model together into a system that provides the ideal characteristics for the efficient processing of XPath queries over XML documents. Transformation algorithms and experimental results have also been shown that prove the feasibility of the invention as a distributed caching system especially tailored for semi-structured data.

A Micro Inertial Measurement Unit (IMU) which is based on MEMS accelerometers and gyro sensors is developed for real-time recognition of human hand motions, especially as used in the context of writing on a surface. Motion is recorded by a rate gyro and an accellerometer and communicated to a Bluetooth module, possibly to a computer which may be 20 to 30 feet or more from the sensor. The motion information generated and communicated is combined with appropriate filtering and transformation algorithms to facilitate a complete Digital Writing System that can be used to record handwriting on any surface, or on no surface at all. The overall size of an IMU can be less than 26 mm×20 mm×20 mm, and may include micro sensors, a processor, and wireless interface components. The Kalman filtering algorithm is preferably used to filter the noise of sensors to allow successful transformance of hand motions into recognizable and recordable English characters. The immediate advantage is the facilitation of a digital interface with both PC and mobile computing devices and perhaps to enable wireless sensing.

The invention relates to the field of image processing and computer vision technologies, in particular to an extraction method for characters in a form document image. The extraction method includes a first step of extracting line segments in the image through edge detection and Hough transformation algorithm, a second step of estimating an inclined angle of the whole image according to direction distribution of the line segments and carrying out inclination correction on the image, a third step of connecting the line segments in the horizontal direction and the perpendicular direction, and locating table cells of a form, a fourth step of carrying out image binaryzation and segmenting a full line of characters in the table cells through a maximum between-cluster variance method, and extracting the characters in the table cells through a window sliding method, and a fifth step of carrying out restoration on deletion of strokes of the characters according to statistics features of frame lines of the table cells. The extraction method is good in flexibility and capable of effectively solving the problems of adhesion between the characters and overlap between the characters and form lines, and greatly reduces the influence of the adhesion and overlap on optical character recognition (OCR).

The invention provides a three-dimension image processing method comprising the following steps: obtaining current image acquisition device position coordinates from a world coordinate system, obtaining a first direction vector and a second direction vector, and using a view transformation algorithm to obtain a transparent matrix; obtaining preset near plane vertex coordinates a near plane distance and a far plane distance from an image acquisition device coordinate system, and using a projection transformation algorithm to obtain a projection matrix; multiplying the transparent matrix with the projection matrix so as to obtain a transformation matrix; multiplying the transformation matrix with texture initial vertex coordinates matched with added augmented reality elements so as to obtaintexture target vertex coordinates matched with the augmented reality elements; using the texture target vertex coordinates to render the augmented reality elements so as to form a three-dimension image. The mobile terminal can rotate to drive the image acquisition device to rotate, so the three-dimension image corresponding to the augmented reality elements can make corresponding rotations, thusincreasing the fusing level between the three-dimension image and a true background image, and improving the image authenticity; the invention also provides a three-dimension image processing device,a storage medium and computer equipment.

The invention discloses a fruit location identification method. The fruit location identification method includes that using a neural network color classification model to divide to obtain a fruit pixel region; using an edge detection algorithm to carry out edge detection on a fruit pixel region image without background to obtain the fruit boundary; carrying out Hough transform circle detection on the fruit edge image through a shape locating Hough transform technology in the image to obtain fruit circle and circle center position, and realizing to divide the overlapped region of the fruit; based on image depth information, camera pose information and world position information, using a world coordinate transformation algorithm to extract the world coordinates of the circle center of each fruit to realize the quick detecting, precise locating and precise counting for the fruit.

A method of text-based authentication that accounts for positional variability of biometric features between captured biometric data samples includes capturing biometric data for a desired biometric type from an individual, and processing the captured biometric data to generate a biometric image and a biometric feature template. A selected conversion algorithm is executed by superimposing a positional relationship medium on the biometric image. The positional relationship medium includes a plurality of cells textually describable with words derivable from the positional relationship medium. The positions of biometric features are permitted to vary in overlapping border regions within the positional relationship medium. The method also includes identifying the position of at least one biometric feature within the overlapping border regions and generating a plurality of words for the at least one biometric feature.

The invention belongs to the technical field of image processing, and discloses an image conversion method and system based on a generative adversarial network and a ResNts technology. The image conversion method comprises the following steps of firstly, designing an enhanced high-resolution dual GAN image conversion algorithm by utilizing a dual learning method combining two GAN confrontation mechanisms and a norm loss function, and realizing image conversion based on unsupervised learning by adopting an unlabeled data set through a model; secondly, realizing the conversion from source distribution to target distribution by introducing constraint conditions for reconstructing a consistency loss function, and then reconstructing the source distribution; and finally, adding a stable normalization layer in the discriminator. Compared with the resolution of images generated by Pix2pixGAN, CycleGAN and DualGAN, the average values of the ERGAN algorithmPSNR/SSIM provided by the invention are improved by 16%/35%, 2%/9% and 4%/6% respectively.

The invention discloses a one-hand operation method of a large-screen touch screen mobile phone. The one-hand operation method comprises the following steps of: step 1, starting a program and starting a touch screen monitoring program; step 2, monitoring sliding operation of a user on a touch screen by a background; step 3, judging whether a thumbnail starting gesture is detected, if the thumbnail starting gesture is detected, carrying out a step 4, otherwise, returning to the step 2; step 4, generating a thumbnail through using a thumbnail generating algorithm; step 5, monitoring touch operation of the user on the thumbnail by the background; step 6, judging whether thumbnail operation is detected, if the thumbnail operation is detected, processing the thumbnail into a corresponding screen touch point by using a coordinate transformation algorithm, so as to enable the program to carry out corresponding screen operation, simultaneously update the display of the thumbnail, otherwise, carrying out a step 7; step 7, by the program, detecting whether closing operation is carried out on the thumbnail, if so, carrying out a step 8, otherwise, returning to the step 5; and step 8, closing the thumbnail, and ending the program. According to the one-hand operation method of the large-screen touch screen mobile phone, the touch control by one hand to the large-screen touch screen mobile phone is enhanced, and the user can enjoy the one-hand operation or the double-hand operation at any time, so that the user experience is obviously improved.

A highly time resolved impedance spectroscopy that enhances the measurement of the dynamics of non-stationary systems with enhanced time resolution. The highly time resolved impedance spectroscopy includes an optimized, frequency rich a.c., or transient, voltage signal is used as the perturbation signal, non-stationary time to frequency transformation algorithms are used when processing the measured time signals of the voltage and current to determine impedance spectra which are localized in time; and the system-characterizing quantities are determined from the impedance spectra using equivalent circuit fitting in a time-resolution-optimized form.

The invention discloses a method for identifying the mode of a switch of a substation based on infrared detection. The method comprises the following steps of: 1, inputting an image to be identified, acquiring a switch area of the image by image registration, and sequentially graying, binarizing and thinning the switch area; 2, linearly detecting the image of the thinned switch area by using a Hough transformation algorithm; and 3, determining a detection result, and identifying the state of the switch by using a switch state determination condition. An H matrix between two images is obtainedby matching of the features of scale invariant feature transform (SIFT), the switch area in the image is found, and then the state of the switch is identified. Experiments show that: by the method, the problem of identification of a power switch can be effectively solved; an important significance is provided for automatic monitoring of the power equipment of an intelligent substation; the burdens on an inspector of the substation can be reduced; and detection efficiency is greatly improved.

A method for loading data into a datamart. The method includes identifying business data in a first format from a real business process and identifying a desired second format. The method includes designing a transformation algorithm that transforms the business data in the first format into the second format and implementing the transformation algorithm in computer executable code. The method includes running the computer executable code on business data in the first format and generating business data in the second format and storing the business data in the second format in a datamart.

The invention realizes a new robot autonomous navigation system based on double single-line laser radars. Compared with mainstream indoor robot navigation systems in the prior art, only by use of two low-cost single-line laser radars as a source of information, navigation and obstacle avoidance effects of the new robot autonomous navigation system are obviously superior to that of schemes with same cost in the prior art. The new robot autonomous navigation system based on the double single-line laser radars includes a double single-line laser radar system, an upper navigation system, a bottom control algorithm and a motion execution system; the upper navigation system includes a SLAM algorithm and a coordinate transformation algorithm; a current space two-dimensional map can be dynamically constructed by use of original laser radar data acquired by horizontally-installed radars by use of the SLAM algorithm, and displacement information of a robot platform can be calculated.

An imaging system and method of application, including lens designs tailored to be used with particular transformation algorithms, electronic hardware and algorithms for image transformations is described. Exemplary application of the system including automotive, photographic and medical endoscopic are also described. The system enables improved image view and allows customization of views by the end user even after installation of the image system hardware.

The invention discloses a convolutional neural network-based character recognition method, and mainly aims at solving the problem that existing manual engineering drawing data input is time-consuming,labor-consuming and high in error rate. The method comprises the following main steps of: 1) converting grayscale images into binary images by adoption of Otsu method-based optimal global threshold value processing; 2) rotating to correct inclined images; 3) extracting a table box by adoption of a mathematic morphological algorithm; 4) obtaining an area where characters are located; 5) rotating inclined character strings for a corresponding angle to a horizontal direction; 6) searching circles in the images by adoption of a circular Hough transformation algorithm, and extracting welding beadnumbers in an engineering drawing; 7) carrying out image-text recognition on the engineering drawing by adoption of a convolutional neural network method; 8) correcting simple errors through an encoding rule and automatically correcting recognition errors; and 9) outputting and storing the drawing data. In image-text recognition of engineering drawings, the method has high correctness and timeliness, and is capable of realizing efficient recording and management of engineering drawing data.

The invention provides a method and a system for transforming a map coordinate system. The method comprises the following steps: taking a coordinate point A in a map range of the current map coordinate system, and setting the coordinate point A as an original point of a two-dimensional coordinate system; transforming the coordinate point A to obtain a coordinate point A' by utilizing a coordinate transformation algorithm, and solving a mirror-image coordinate point B of the coordinate point A' about a symmetric axis of y=-x, wherein y is a longitudinal axis of the two-dimensional coordinate system, and x is a transverse axis of the two-dimensional coordinate system; transforming the mirror-image coordinate point B to obtain a transformed coordinate point B' according to the coordinate transformation algorithm; transforming the map coordinate system according to a longitude and latitude corresponding relation of the coordinate point A and the transformed coordinate point B'. By adopting the method and the system, the longitude and latitude among different geological coordinate systems can be mutually transformed freely, an effective and feasible corrective method can be provided for the coordinate encryption algorithm, the longitude and the latitude among different SDK (software development kit) of different maps can be corrected, and the precision and working efficiency for collecting the map information can be improved.

Apparatus for radiation therapy combines a patient table, an MRI and a radiation treatment apparatus mounted in a common treatment room with the MR magnet movable through a radiation shielded door to an imaging position. An initial MR image and an initial X-ray image is used to generate an RT program for the patient to be carried out in a plurality of separate treatment steps. Before carrying out the procedure, a registration step is performed using a phantom by which X-ray images are registered relative to MR images to generate a transformation algorithm required to align the MR images of the part of the patient relative to the X-ray images. Prior to each separate treatment step a current MR image of the part of the patient is obtained and the transformation algorithm data is used from the current MR image is used in guiding the RT treatment step.

Methods, systems and software for determine an optical surface model for an optical tissue system using Fourier transformation algorithms. A method of reconstructing optical tissues of an eye comprises transmitting an image through the optical tissues of the eye. The surface gradients from the transmitted image are measured across the optical tissues of the eye. A Fourier transform algorithm is applied to the surface gradients to reconstruct an optical surface model that corresponds to the optical tissues of the eye.

In a document analysis system that receives and processes jobs from a plurality of users, in which each job may contain multiple electronic documents, to extract data from the electronic documents, a method of automatically pre-processing each received electronic document using a plurality of image transformation algorithms to improve subsequent data extraction from said document is provided. The method includes: electronically partitioning each received electronic document page into pieces; automatically processing each piece of the received electronic document page using each of a plurality of image pre-processing algorithms to produce a plurality of image variations of each piece; and analyzing the outputs of subsequent processing and data extraction, on each of the image variations of the pieces to determine which output is best, from the plurality of outputs for each piece.

The invention relates to an image synthesis method based on an image block template and capable of adding modification materials. The steps of the image synthesis method are as follows: 1) a blank canvas is created, a laminated material data buffer cache is created, laminated material data are added, background materials in the laminated material data are drew with colors as the background of the canvas, and the background materials are stored in the laminated material data buffer cache; 2) vector data of the image block template are uploaded and transformed into image blocks in the canvas through spatial coordinates transformation algorithm, and the image blocks are enabled to be displayed completely on the canvas; 3) data flow of an image is obtained and synthesized into one image block of the canvas; 4) the step 3) are repeated until all the synthesis of all images is completed. The image synthesis method is based on the image block template, easy to operate, unlimited in distribution and capable of adding words, images and backgrounds. The adding of postal materials of characteristic images, words and the like is met, and the background is more vivid and personalized.

The invention discloses an electrocardiosignal denoising method based on morphological filtering and lifting wavelet transformation. Firstly, three times of decomposing are conducted on electrocardiosignals f based on the lifting wavelet theory, three layers of high-frequency coefficients and three layers of low-frequency coefficients are obtained, a lifting threshold denoising method is used for processing the high-frequency coefficients, the high-frequency coefficients on the bottom layer and the low-frequency coefficients are reconstructed twice, the reconstructed low-frequency coefficients can be obtained, morphological filtering is conducted on the reconstructed low-frequency coefficients, signal reconstruction is conducted according to the processed reconstructed low-frequency coefficients and the processed high-frequency coefficients on the highest layer, and denoised electrocardiosignals f' are obtained. The electrocardiosignal denoising method has the advantages that the method is simple and easy to achieve, the morphological algorithm and the lifting wavelet transformation algorithm are organically combined, compared with a traditional wavelet algorithm, the high-frequency electrocardio-noise and the low-frequency electrocardio-noise can be removed at the same time, the quality of the denoised signals is improved, computing is easy, the occupied space is small, and the electrocardiosignal denoising method can be achieved on hardware more easily.

A mapping application relies on a modified affine transformation algorithm to provide near real-time navigation of imagery, track-up capability and integration of segment and picture data. The affine transformation is modified to map a “world space” comprised of a plurality of pixel maps onto a view port. The pixel data for the maps is preferably stored in vertical strips of memory to reduce the likelihood of cache misses. The mapping application is most useful for rendering pixel imagery on a platform with limited processing power, limited memory and small display sizes, e.g., “hand held devices” such as a GPS handset, or Pocket PC.