38 results about "Fuzzy filtering" patented technology

A method reduces artifacts in an input image. A variance image is generated from the input image. The input image is partitioned into a plurality of blocks of pixels. A set of classifications is defined. The set of classifications includes smooth, texture, and edge. A particular classification is assigned to each block of pixels of the input image according to the variance image, to generate smooth blocks, texture blocks, and edge blocks. A fuzzy filter is applied to each pixel of only each edge block.

A set of volatile markers are determined which are characteristic of a particular condition or disease, and which will be found in the exhaled breath of a person or odor from other parts of a body or from an entity. These markers are detected in the breath odor or gaseous emanations from the body or entity noninvasively using a volatile substance detector of sufficient sensitivity, such as an artificial olfactory system. The detected marker data is processed in an artificial neural network / fuzzy filter system with an algorithm that intelligently adapts to the individual body or entity and also optionally (if necessary) with a correction algorithm to eliminate environmental and other erroneous contributions to the markers. Any number of markers may be used, depending on how well they correlate with the condition and how accurate a result is desired, i.e. general screening or accurate diagnosis and monitoring.

An invention provides a system and method for filtering pixels in an image using only fixed-point and summation operations. First, a filtering window is centered on an input pixel. Based on a difference between the intensity of the input pixel and its neighboring pixels, fuzzy filter weights are obtained. A sum of the fuzzy filter weights is used to determine a normalization factor. Then, the pixel intensities, fuzzy filter weights and the normalization factor are used to obtain an output pixel corresponding to the input pixel.

A method reduces artifacts in an input image. A variance image is generated from the input image. The input image is partitioned into a plurality of blocks of pixels. A set of classifications is defined. The set of classifications includes smooth, texture, and edge. A particular classification is assigned to each block of pixels of the input image according to the variance image, to generate smooth blocks, texture blocks, and edge blocks. A fuzzy filter is applied to each pixel of only each edge block.

A method filters pixels in a sequence of images. Each image in the sequence is partitioned into blocks of pixels, and the images are processed sequentially. The energy is determined for each block of pixels in each image. The energy of each block is based on variances of intensities of the pixels in the sequence of images. A 3D fuzzy filter is applied to each current pixel in each current block during the sequential processing. The 3D fuzzy filter considers the energy of the block, and the intensities of pixels spatially adjacent and temporally adjacent to the current pixel to remove blocking and ringing artifacts.

The invention relates to an algorithm of a progress bar of a breathing lamp effect. The algorithm includes the steps that firstly, a new control class BurnProgressView is user-defined on the basis of layout control Linearlayout, and an initialization method is added to a construction function; secondly, in onSizeChanged callback, the size of the circular progress bar is initialized according to the control size, and an inner ring area and an outer ring area are determined according to the width of a drawing pen so that gradient radian lines can be conveniently drawn on the two sides of the progress bar; thirdly, a thread of a breathing lamp change is started, the radiuses of a drawing pen fuzzy filter are circularly switched, and a height mark of corrugated lines is changed to be consistent with the rhythm of a breathing lamp; fourthly, in onDraw callback, progresses and effects are sequentially and dynamically drawn. According to the algorithm, the ring-shaped progress bar and the surrounding of the breathing lamp effect are adopted, attractive, personalized, visually and easily understood progress bar display is achieved through the circular progress bar and the breathing lamp effect, the algorithm efficiency is high, few resources are occupied, the universality is good, and the algorithm is suitable for being used in the Android system.

The invention discloses a webpage data capturing and filtering method, which comprises the following steps: a) obtaining webpage Html codes; b) converting the webpage Html codes into an XML (extensive makeup language) sequence table; and c) performing fuzzy matching on the XML sequence table by using information key words to obtain webpage data. According to the webpage data capturing and filtering method provided by the invention, the webpage Html codes are comprehensively subjected to XML serialization at first, and the webpage data in XML files are obtained in a fuzzy filtering way, so that massive amounts of webpage data can be captured and filtered quickly and more accurately.

The invention discloses an image denoising method and system based on threshold value block matching. The image denoising method includes acquiring several images to be processed in time succession, calculating threshold values matched with blocks according to brightness of the images to be processed and processing by fuzzy filter to obtain fuzzy images, layering according to calculation distance of SAD (sum of absolute difference), and matching the blocks of the fuzzy images according to layering search; using the first image of the continuous images to be processed as a reference image, accumulatively processing the images to be processed one by one, and accumulating the image blocks with SAD smaller than the threshold values to the corresponding positions of the reference image so as to obtain cumulative sum and the images; dividing color values of various pixel points in the cumulative sum and the images from the cumulative frequency of the pixel points correspondingly to obtain denoised images. Accordingly, block matching efficiency is greatly improved, accuracy in matching is ensured, denoising effect is improved, and the image denoising method can be conveniently introduced into the industrial production.

The invention provides a filtering method, a device and electronic equipment, and relates to the technical field of filtering. The filtering method comprises the following steps: establishing an overall interval II-type T-S fuzzy time delay system model; establishing an overall interval II-type fuzzy filter model; establishing a filtering error model according to the interval II-type T-S fuzzy time delay system model and the interval II-type fuzzy filter model, and filtering according to the filtering error model. The filtering method, device and electronic equipment provided by the invention can better process a nonlinear system with parameter uncertainty, and accomplish the design of the filter parameter, so that the filtering is more stable.

The invention provides a two-joint robot tracking fuzzy control design method, and relates to the technical field of robot control. The method includes that a two-joint robot system is built and has two permanent magnet direct current motors to control a first joint and a second joint to rotate respectively, wherein the first joint is hinged to the second joint. According to the dynamics principle, a mathematical model of the two-joint robot system is built, and the mathematical model belongs to a nonlinear system with strong coupling. The nonlinear system is converted into a fuzzy system based on the mathematical model and the expression method of the fuzzy model. A tracking fuzzy output feedback controller is designed based on measurement signals of the two-joint robot system. Finally, adelta operator fuzzy filter is designed for filtering measurement noise. Measurement noise existing in the plane position measurement at the tail end of the robot joint can be effectively eliminated,and the control precision of the tail end position of the two-joint robot is guaranteed.

The embodiment of the invention provides an image fuzzy processing method and device. The method comprises that a boundary of a clear area and a fuzzy area in an image is arranged; a sampling fuzzy parameter of the fuzzy area is arranged according to the size of the image to be processed; sampling fuzzy processing is performed on the fuzzy area by adopting the sampling fuzzy parameter of the fuzzy area via arrangement; fuzzy templates corresponding to pixel points in the fuzzy area are calculated; and fuzzy filtering is performed on all the pixel points in the fuzzy area after sampling fuzzy processing by adopting the obtained fuzzy templates. With application of the method and the corresponding device, smaller fixed templates can be used so that fuzzy processing speed is enhanced and adaptability is high.

The present invention provides a generation method of an image bokeh effect and an image acquisition device. The generation method of the image bokeh effect comprises the following steps: obtaining image frames including a current input pixel, and performing fuzzification processing of the image frames by respectively employing a first image fuzzy filter and a second image fuzzy filter to generate a plurality of first blurred images and second blurred images with respect to different fuzzy degrees; calculating the distance between the current input pixel and a focusing plane to obtain a current distance; respectively selecting a first current blurred image and a second current blurred image from the first blurred images and the second blurred images according to the current distance, and combining the first current blurred pixel of the first current blurred image and the second current blurred pixel of the second current blurred image according to the current distance to generate a current output pixel.

The image processing method and device is used to convert progressive scanned and displayed image data into interlaced scanned and displayed image data. The device includes a fuzzy filter lens unit, a scanning size regulator and a converter. The method includes the following steps: receiving progressive scanned and displayed image data; substituting the color value of each pixel with a combined value of the color values of the said pixel and adjacent pixels so as to obtain fuzzy filter lens processed progressive scanned and displayed image data; regulating the picture size to obtain new progressive scanned and displayed image data; and converting the new progressive scanned and displayed image data into interlaced scanned and displayed image data for output.

The invention discloses a target detection and identification method under rain and snow weather conditions, which comprises the following steps: firstly, acquiring to-be-detected rain and snow weather background image data containing a target, and constructing a training data set under the corresponding rain and snow weather conditions through a fuzzy filter; optimizing a separation model of a rain and snow layer and a background layer, and respectively processing a rain area and a rain-free area so as to weaken detail loss of the rain and snow-free area; building a context expansion rain removal network based on the scene information so as to restore background images under different rain and snow degrees; building a convolutional neural network based on local feature learning, and improving the target recognition rate; taking the rain and snow removed image as the input of an optimized target detection network model, extracting an interested area in the image data by the target detection network model, and outputting a target category; the rain and snow removal target detection model provided by the invention has good generalization and universality, and can be widely applied to high-precision automatic identification and detection of various targets under different rain and snow degrees and other actual scenes.

The invention discloses an image processing method, a device, a storage medium and an electronic device. The method comprises the following steps of: acquiring an original image to be processed; obtaining The target image by flooding the pixels of the original image based on the mean value fuzzy filter. The invention achieves the effect of improving the speed of image processing.

The invention discloses a laser radar echo data processing method for improving fuzzy neural network filtering, and the method comprises the steps: carrying out the modeling of a noise-containing laser radar echo signal, constructing a simulation ideal signal model, and correcting the echo model through improving an atmospheric pulse laser radar equation; secondly, updating a weight parameter of a filtering algorithm system through an LMS (Least Mean Square) algorithm; secondly, designing a self-correcting fuzzy neural network system and simultaneously processing numerical data and language information of laser radar echo signals; and finally, utilizing a fuzzy filtering-X LMS algorithm to update filtering system parameters, and the flight time, and obtainingthe pulse width and the amplitude by checking the signal quality through a signal-to-noise ratio. According to the method, noise can be better eliminated, and the problem of low convergence is solved, so that the performance of a laser radar echo signal data processing system is enhanced.

The image processing method and device is used to convert progressive scanned and displayed image data into interlaced scanned and displayed image data. The device includes a fuzzy filter lens unit, a scanning size regulator and a converter. The method includes the following steps: receiving progressive scanned and displayed image data; substituting the color value of each pixel with a combined value of the color values of the said pixel and adjacent pixels so as to obtain fuzzy filter lens processed progressive scanned and displayed image data; regulating the picture size to obtain new progressive scanned and displayed image data; and converting the new progressive scanned and displayed image data into interlaced scanned and displayed image data for output.

The invention discloses a laser flight detection method and implementation for the sag of an electric power overhead cable, and the method comprises the steps: enabling a rotor aircraft to fly to a position above a detected overhead cable through an airborne laser radar / distance measuring sensor of the rotor aircraft, and carrying out the measurement; then screening measurement data, wherein the measurement data are used for calculating the sag of the aerial cable and the distance to the ground; 1) the measuring method for a certain measuring point of any cable includes the steps that the rotorcraft keeps the flight height and the flight speed unchanged, and transversely sweeps the cable at a constant speed in the flight detection plane in the direction perpendicular to the trend of the cable; meanwhile, the laser radar / distance measuring sensor emits a detection laser beam sequence which is perpendicular to the flight detection plane and faces downwards, and cable and obstacle height data below the flight detection plane are continuously obtained; and 2) measurement data is screened by using a fuzzy filter to obtain effective data.

The invention provides a high-frequency pantograph-catenary dynamic response signal filtering method and device, and the method comprises the steps: carrying out the Fourier transform and inverse Fourier transform, and obtaining a high-frequency pantograph-catenary dynamic response signal after high-pass filtering; filtering an interference signal corresponding to the single-point high-frequency abnormal value by using a classification regression decision tree CART according to the signal change characteristics after high-pass filtering to obtain a first signal; according to the relative peakvalue, filtering an interference signal corresponding to the section abnormal value by using a fuzzy filter to obtain a second signal; and according to the low-frequency trend information, filtering out the interference signal corresponding to the fragment abnormal value by using a wavelet packet to obtain a signal after the complete interference signal is removed. Different interference signals are filtered by different methods, so that normal signals are not damaged, and interference signals in high-frequency pantograph-catenary dynamic response signals are completely and accurately filtered.

The invention discloses a Beidou deformation monitoring and positioning method based on fuzzy confidence filtering. The method comprises the following steps of: firstly, dividing a monitoring period into a plurality of sub-periods, and calculating a positioning result of each sub-period; then according to satellite distribution, error sources and a satellite epoch number of each sub-period, adopting a fuzzy set theory to determine the fuzzy confidence of each positioning result; and finally, utilizing the fuzzy confidence to carry out weighted filtering so as to obtain a final positioning result of the whole monitoring period. According to the invention, the positioning calculation process is detailed, each factor influencing the positioning precision is fully considered, and the final positioning result is enabled to achieve a higher positioning precision; and especially when the weather changes in the monitoring period, the method has obvious advantages in positioning precision, so that the method has a wide application prospect.

The invention relates to a trust evaluation method based on fuzzy filtration. According to the method, based on the Dirichlet reputation evaluation system and the fuzzy theory, an evaluation matrix is established to refine user evaluation, calculate a direct trust vector, conduct evaluation filtration, conduct evaluation convergence, calculate a recommended trust vector and calculate a comprehensive trust value, and finally a transaction is conducted between a solicited node and a corresponding node which has a high comprehensive trust value; in this way, the technical problem of the prior art that the reputation value of a node in a P2P network is too low, and therefore the reputation value is whitewashed or driven up maliciously is solved. Meanwhile, by updating the Dirichlet prior basic rate, the motivation of providing high-quality services of the nodes is effectively improved, the node whitewashed probability is reduced to a certain degree, and the stability of a trust model is improved.

The invention provides a facial physiological information shielding method without changing a video effect belongs to the field of personal information safety, and the gradual maturation of a video-based non-contact physiological signal detection method makes it possible to obtain physiological parameter information such as personal heart rate, heart rate variability and the like only through a common video. Without knowing, personal physiological information may leak and be analyzed and utilized along with the propagation of the video. According to the facial physiological information shielding method, the video effect is not changed, non-contact physiological signal detection is interfered through single-channel periodic fuzzy filtering and single-channel periodic noise adding technologies, the rPPG principle and the detection process are fully considered, and traditional physiological signal detection methods such as ICA and CHROM can be rapidly and efficiently resisted.

The invention belongs to the technical field of image super-resolution reconstruction, and particularly relates to a dual-energy CT (Computed Tomography) super-resolution image reconstruction method and a dual-energy CT super-resolution image reconstruction system based on sparse representation, which are characterized in that paired high-resolution and low-resolution dual-energy CT images are obtained through de-downsampling operation of an original truth value image fuzzy filter, and paired matched image segmentation blocks are extracted from CT images in a training set; in order to obtain similarity of sparse representation, two pairs of dictionaries are jointly trained for the high-energy-spectrum CT image and the low-energy-spectrum CT image respectively, and each pair of dictionaries comprises dictionaries of low-resolution image blocks and dictionaries of high-resolution image blocks; the sparse representation coefficients extracted from the low-resolution input block are then multiplied by the corresponding high-resolution dictionary for reconstructing the high-resolution output. According to the method, better image quality and clearer visual effect can be obtained by reconstructing the high-resolution image in combination with a proper amount of iterative operation, the resolution of the dual-energy CT image can be effectively improved according to the image quality evaluation index and the peak signal-to-noise ratio, and the method is convenient for actual scene application.

The invention discloses a webpage data capturing and filtering method, which comprises the following steps: a) obtaining webpage Html codes; b) converting the webpage Html codes into an XML (extensive makeup language) sequence table; and c) performing fuzzy matching on the XML sequence table by using information key words to obtain webpage data. According to the webpage data capturing and filtering method provided by the invention, the webpage Html codes are comprehensively subjected to XML serialization at first, and the webpage data in XML files are obtained in a fuzzy filtering way, so that massive amounts of webpage data can be captured and filtered quickly and more accurately.

The invention discloses an information acquisition method for English pronunciation. The information acquisition method comprises the following steps of S1, acquiring an audio signal and amplifying the audio signal; S2, performing analog filtering on the amplified audio signal; S3, converting the analog filtered signal into a digital signal, and extracting audio characteristic parameters of the digital audio signal, such as sound generation time, frequency spectrum centroid, frequency spectrum flux, fundamental tone frequency, sharpness and the like; S4, the audio characteristic parameters being matched with a sound source model in a standard sound source database, then the digital audio signals being matched with syllables and phonemes in the sound source model to obtain matching degrees, and phoneme correction being carried out according to the difference of the matching degrees; S5, combining the corrected phonemes into the digital audio signal; and S6, performing fuzzy filtering on the synthesized digital audio signal, and outputting an audio signal.

A computer graphics processor with a forward mapped renderer is provided. The renderer includes a texture space rasterizer (TS) that rasterizes primitives in texture space, a color generation unit that determines the color of the output of the texture space rasterizer (TS) and provides color samples along with coordinates PS), a two-pass screen space resampler (SSR1, SSR2) for resampling the color samples determined by the color generation unit (PS), and at least one pass correlation with said screen space resampler (SSR1, SSR2) There is at least one connected one-dimensional fuzzy filtering unit (1PB, 2PB), which performs one-dimensional fuzzy filtering before performing the at least one pass.

The invention relates to a method for designing an event trigger filter of an cyber-physical system under spoofing attack, and the method comprises the following steps: designing an event trigger mechanism according to a formulated T-S fuzzy model, and constructing measurement output of a filter to describe the spoofing attack suffered by the cyber-physical system; according to the spoofing attack suffered by the information physical system, constructing an elastic fuzzy filter corresponding to the spoofing attack; according to the elastic fuzzy filter, constructing a filtering error system for detecting the filtering effect of the elastic fuzzy filter; and under a spoofing attack, verifying the stability of the information physical system in which the elastic fuzzy filter is constructed. The invention aims to improve the bandwidth utilization rate of the information physical system and reduce the false triggering of the data packet.

The invention belongs to the technical field of endoscope image processing, and provides an endoscope image processing method and system, a storage medium and equipment. In order to solve the problem that low-quality images occupy the storage capacity of a database and bring workload to image auditing personnel, the endoscopic image processing method comprises the steps that after an image acquisition instruction is received, a current endoscopic image is stored as a first recognition frame, and a plurality of endoscopic images with the similarity with the first recognition frame exceeding a first preset threshold value are continuously acquired; performing fuzzy filtering processing on the acquired endoscopic image; scoring the filtered endoscope images based on a preset image quality evaluation model, and selecting the endoscope image with the highest scoring quality as a second recognition frame; and comparing the score quality of the first identification frame and the score quality of the second identification frame based on a preset image quality evaluation model, and only reserving one of the two frames with the best quality as an endoscopic image acquired by the current image acquisition instruction and storing the endoscopic image. The workload of image auditing personnel can be reduced after the number of acquired images is reduced.

The embodiment of the invention provides a community information management method, and the method comprises the steps: carrying out the live-action shooting in a to-be-managed community through a mobile device, obtaining the shot live-action image information, carrying out the photogrammetry, and carrying out the fuzzy filtering in a three-dimensional live-action model relation database, performing first-level object recognition on the live-action image by utilizing a constructed object learning model, associating the first-level object recognition with object information stored in a three-dimensional live-action model relational database, and recognizing a coupling object on the surface of the first-level object by utilizing a coupling object recognition model corresponding to the recognized object, and determining a pixel corresponding relation between each floor resident and the live-action image based on the coupling object. The pixel corresponding relation between each floor resident and the real scene image is determined based on the coupling object, so that the identification of the building in the community and the object information on the surface can be accurate to the home level, the error is reduced, and the accuracy is improved.