36results about How to "No resolution reduction" patented technology

The invention relates to a post-processing method for removing ring artifacts in a CT (computed tomography) image, and the post-processing method comprises the following steps: converting the CT image with ring artifacts from a rectangular coordinate image to a polar coordinate image; extracting the ring image in polar coordinates with a comb filter in a polar coordinate image domain; converting the ring image extracted out of the polar coordinates to a ring image in rectangular coordinates; and subtracting the ring image in the rectangular coordinates from the original CT image with the ringartifacts to obtain a ring-free image. Through frequency domain comb filtering technology, the post-processing method provided by the invention subtracts the corresponding ring artifact image from the original image, without reduction of image resolution and change of quality of the original image. The post-processing method achieves the processing effect, and meanwhile the speed meets product requirements and has space for further optimization, so that the post-processing method is fully suitable for a CT machine with higher requirements for the reconstruction speed.

The invention relates to a gamma ray spectrometry radionuclide identification method utilizing multiple detectors. The gamma ray spectrometry radionuclide identification method comprises the following steps: step 1) respectively seeking peaks of spectra acquired by multiple detectors from substances at the same time; step 2) utilizing the peaks sought in the step 1) to constitute merged peaks, wherein each merged peak is merged by a plurality of the peaks with energy difference in a preset range; and step 3) performing radionuclide identification by utilizing the merged peaks in the step 2). According to the method disclosed by the invention, high sensitivity and omnidirectional detection which are achieved by the multiple detectors can be utilized, and the problem that resolution is reduced due to addition of the multiple spectra can be also avoided. The method is not limited by the number of the detectors and can be used even part of the detectors is damaged.

The invention discloses an improved RDA-based method for inhibiting a synthetic aperture radar image target side lobe. In order to obtain a super-resolution SAR image, the method comprises the following specific processing steps of azimuth direction Fourier transformation of a two-dimensional echo signal, complementary migration correction, distance direction Fourier transformation, consistent migration correction and distance compression, distance direction inverse fast Fourier transformation, additional phase correction, azimuth direction inverse Fourier transformation and azimuth compression of a fusion inversion matrix. Compared with conventional imaging processing modes of matched filtering and the like, the processed imaging result energy is more concentrated, the space resolution ratio is significantly improved, and the super resolution target is achieved. Compared with a complete matrix-based SAR data processing, the method disclosed by the invention is combined with a high-efficiency traditional imaging algorithm, so that the processing efficiency is significantly improved.

The invention discloses a rapid microscopic section scanning method with image space compensation motion and an apparatus thereof. The method comprises the following steps that: a section is divided into several continuous scanning strips longitudinally; a width of the each scanning strip is set as a dimension of corresponding object space field of an area array image sensor at a longitudinal direction; the each scanning strip is divided into several continuous scanning fields transversely; the length of the each scanning field is set as the dimension of the corresponding object space field of the area array image sensor at a transverse direction. During working, the each scanning strip of the section is controlled to successively pass through the object space field corresponding to the area array image sensor at a uniform speed. And when passing through one scanning field, the area array image sensor is exposed once. While exposing, the area array image sensor moves at an imaging movement speed (including the direction and magnitude) of the section on the area array image sensor so as to eliminate influence of image motion. The scanning speed can be effectively increased without resolution reduction.

The invention provides a railway tunnel crack detection method based on an improved residual network. The method comprises the steps: collecting a railway tunnel crack image, and obtaining a railway tunnel crack detection data set; the method also comprises the following steps: carrying out enhancement processing on images in the railway tunnel crack detection data set and randomly dividing the images into a training set and a test set; improving the structure of the residual network, and training the improved residual network by using the images in the training set; and carrying out crack classification detection on the railway tunnel image by using the trained residual network structure. According to the method, a ResNet residual network structure is improved to form a pyramid cavity convolution module; a pyramid cavity convolution module is added to improve the underlying receptive field of the ResNet network and improve the classification accuracy; a combined loss function based on metric learning is adopted to distinguish similar differences between different classes, the omission ratio and the false drop rate of cracks are reduced, and therefore detection of small cracks under a complex background is better achieved.

The invention relates to a 3D display substrate which comprises a plurality of sub pixels, a driving unit and a plurality of planoconvex lenses positioned at the light emitting sides of the pluralityof sub pixels, wherein the plurality of sub pixels are arranged in a one-to-one correspondence way to the plurality of planoconvex lenses; the light emitting surface of each sub pixel is positioned onthe focal plane of the corresponding planoconvex lens; each sub pixel comprises a plurality of son pixels arranged in an array manner; and the driving unit is used for driving the plurality of son pixels to display images at different gray scales so as to form a 3D display image. The invention also relates to a 3D display device and a display method.

The invention discloses a display device and a display method. The display device comprises a control unit and two display units, wherein the control unit comprises a first processing module and an adjustment module; the first processing module receives first display data and transmits the first display data to the adjustment module and a first display unit, and the first display unit displays the first display data; the adjustment module adjusts the display brightness and the display time of the first display data to obtain second display data, and a second display unit displays the second display data; the first display unit and the second display unit are provided with display pixel points which are in one-to-one correspondence, and the distance between every pair of corresponding display pixel points is L1; the distance between the display pixel points of the first display unit and the second display unit is L0; and L1 is less than L0. The display method comprises corresponding steps. Based on the fact that the resolution capacity of human eyes on color details is far lower than the resolution capacity of human eyes on brightness details, the invention can realize the dual display of an image.

The invention discloses a method and a device for correcting a ring artifact. Polar coordinate conversion, related total variation processing, mid-value extraction and iterative correction are used for achieving iterative ring artifact correction. The method specifically comprises the steps of smoothening edge protection of an image by using the related total variation, thus acquiring detail and artifact images, and applying the images to the correction of the ring artifact; and continuously extracting the ring artifact by using the iterative correction method, thus achieving a ring artifact correction result of the image. According to the method and the device provided by the invention, during a process of removing the ring artifact, details of the original image are not damaged, and an image resolution ratio is not reduced, and a processing flow is simple; in addition, the ring artifact is only extracted in an image domain, so that the process is completely compatible with a clinic using process. The method and the device are more suitable for clinic application.

The invention discloses a parallel multi-scale attention mechanism semantic segmentation method based on deep learning, and the method comprises the steps: firstly carrying out the preprocessing of animage data set, and improving the segmentation precision and robustness of a model; based on ResNet50 obtained after adjustment of a fifth convolution layer, aggregating multi-scale semantic information through a parallel multi-scale attention module connected to the top of a base network. And finally, recovering the image size through bilinear up-sampling. According to the method, a similarity EDS module is added behind a feature map obtained by five original parallel convolution kernels with different sizes. By adding the attention mechanism, important semantic information in a feature mapobtained by five parallel expansion convolution is enhanced, and secondary semantic information is inhibited.

The invention discloses a display apparatus and a sub pixel rendering method. The display apparatus comprises multiple pixel groups, and the pixel groups comprise first sub pixel groups and second sub pixel groups; the first sub pixel groups comprise first sub pixels, second sub pixels and third sub pixels; the second sub pixel groups comprise first sub pixels, fourth sub pixels and fifth sub pixels; the second sub pixels and the fifth sub pixels are the same in color type yet different in chroma; and the third sub pixels and the fourth sub pixels are the same in color type yet different in chroma. According to the invention, the sub pixels with the same color type yet different chroma are employed in sub pixel arrangement so that the penetrating power of a display panel is improved while the resolution is not reduced and the color gamut is not broadened.

The invention provides a scale domain multichannel filtering method and device based on seismic signal continuous wavelet transformation. The method includes the following steps that seismic signals are decomposed into multi-scale-domain characteristic signals through continuous wavelet transformation; scale-space F-X domain multichannel prediction filter processing is conducted on the multi-scale-domain characteristic signals; the seismic signals are reconstructed according to the multi-scale-domain characteristic signals on which scale-space F-X domain multichannel prediction filter processing is conducted. According to the method and device, the seismic signals are decomposed into the multi-scale-domain characteristic signals, random noise of seismic sections of different scale is selectively eliminated according to the characteristics of different signals and noise and through F-X domain multichannel prediction filtering, the seismic signals are reconstructed on the basis, and therefore interference waves can be effectively eliminated without reducing resolution of signals, the effect of increasing the signal to noise ratio of effective signals is achieved, and effective signals with weak energy can be extracted from complex seismic signals.

The invention discloses an attention generative adversarial network-based unsupervised image rain removal method, relates to the field of computer vision, and mainly relates to a neural network model capable of carrying out unsupervised learning and improving a picture rain removal effect. The problem that paired data are difficult to acquire when the generative adversarial network is trained can be effectively solved, and an attention mechanism is introduced, so that the network focuses on a rain area when processing an image, and a more ideal rain-free image is output. The method comprises the steps of 1, constructing a data set; 2, building a convolutional neural network; 3, training the network; 4, putting the process into actual use. Two discriminators of the original cyclic adversarial generative network are replaced by one discriminator, the network is simplified, the calculation amount is reduced, the discrimination capability of the discriminator is improved, and the image recovery capability of the generator is further improved. Therefore, the network focuses on a rain area when processing the image, and a more ideal rain-free image is output.

Embodiments of the invention provide a display device and a control method thereof, which relates to the field of display technology, can support 3D display, and does not reduce the resolution when 3D display is realized. The display device comprises a display unit, a light adjustment structure, and a controller. The display unit is used to display n images of different view angles periodically on a time-division basis, wherein one frame displays an image of a viewing angle, n is not less than 2, and n is a positive integer. The light adjustment structure can be transformed into m different states and is used in the different states to adjust the light emitted by the display unit to directions toward different viewing positions, wherein m is greater than n and is a positive integer. The controller is connected with the light adjustment structure and is used to control the light adjustment structure to change to n different states periodically to transmit images of n different view angles to n viewing positions. The display device and the control method are used to realize 3D display without reducing the resolution.

The invention discloses a cloud computer system, a coding method and device, equipment and a storage medium, and the method comprises the steps: detecting whether the utilization rate of a local resource meets a preset utilization rate condition or not, and determining a first preset coding format as a target coding format if the utilization rate does not meet the preset utilization rate condition; if so, judging whether the decoding capability of the client meets a decoding condition corresponding to a second preset coding format, and if so, determining the second preset coding format as a target coding format; and encoding by using the target encoding format, and sending the encoded data to the client, so that the client decodes the encoded data. According to the invention, the coding format is dynamically switched according to the decoding capability of the client and the real-time resource utilization rate of the host, and the overall performance of image display is improved underthe condition that the current resolution, frame rate and delay performance of the host and the client are not reduced.

The invention discloses a bionic visual sensor with an enhanced end side contour, and the sensor comprises a sensor main body element module which comprises one or more pixels, each pixel comprises a plurality of elements which are used for converting an optical signal into an electric signal, and the elements in each pixel have a light intensity information exchange capability, the sensitivity of the elements is regulated and controlled through light intensity information exchange, the light intensity difference between the elements is amplified, and the purpose of outline enhancement is achieved. The visual sensor is based on a lateral inhibition function of horizontal cells in a human retina, and on the basis of image acquisition, end-side contour enhancement and image feature extraction can be achieved on the hardware level through an antagonism effect between elements in pixels, so that information redundancy is reduced, the imaging speed is increased, and the imaging precision is improved. And self-adaption and self-regulation are realized, and manual regulation is not needed.

A drawing apparatus of the present invention includes a function of a mark position detecting apparatus for detecting a position of a mark on a substrate held by a workbench. The drawing apparatus acquires a first image (83a) representing a first region including a position on a design of a mark on a substrate and two second images (83b) representing a second region that only partially overlaps the first region while the workbench is moving. The first image (83a) and the second images (83b) are synthesized to generate a synthesized image (85). The drawing apparatus finds the position of the marker on the basis of the synthesized image (85). As a result, it is possible to easily expand the field of view without reducing the resolution of an imaging unit, and to reduce detection errors in the position of the mark on the substrate.

The invention discloses an amorphous silicon flat panel detector, comprising a photosensitive layer, wherein the photosensitive layer is composed of a photodiode circuit made of amorphous silicon; a photosensitive surface is formed on the surface of the photosensitive layer; the photosensitive surface is in the shape of a rectangle, and the length of each side of the rectangle is no less than 4 cm; the photosensitive surface is covered with a transparent film for closely bonding with a self-luminous object; the self-luminous object is a film used in Western blot or Southern blot detection; thefilm contains to-be-detected protein or nucleic acid; the protein or nucleic acid binds to a specific antibody or nucleic acid; the specific antibody or nucleic acid is cross-linked with horseradishperoxidase; in virtue of catalysis by hydrogen peroxide, the chemiluminescent substrate of the horseradish peroxidase is allowed to give out fluorescent light; and the fluorescent light can penetratethe transparent film and is then captured by the photosensitive layer to form an image. The amorphous silicon flat panel detector of the invention can meet requirements on sensitivity and resolution and has size advantages and obvious cost advantages during large-format application.

The invention discloses a three-dimensional endoscope system compatible with two-dimensional endoscopes. The three-dimensional endoscope system comprises a first object lens, a second object lens, a first image transmission column, a second image transmission column, a first image diaphragm, a second image diaphragm, a first focusing lens group, a second focusing lens group, a first deflection prism, a second deflection prism, a first cylindrical lens and a second cylindrical lens. The three-dimensional endoscope system has the advantages that the first deflection prism is used for reflectingfirst light beams which are received by the first focusing lens group and are projected from the first focusing lens group, the second deflection prism is used for reflecting second light beams whichare received by the first focusing lens group and are projected from the second focusing lens group, accordingly, the distances between the optical axes of the first light beams and the second light beams can be changed, and the three-dimensional endoscope system can be matched with image sensors of two-dimensional endoscope systems; the first cylindrical lens is used for compressing the first light beams reflected from the first deflection prism, the second cylindrical lens is used for compressing the second light beams reflected from the second deflection prism, and accordingly flattened first images and flattened second images can be acquired by first light sensing regions and second light sensing regions of the image sensors.

The invention discloses a gas ultrasonic flow sensor. The gas ultrasonic flow sensor comprises a shell, a piezoelectric unit, a matching piece, a lining plate and a potting adhesive, wherein the matching piece is fixedly arranged at the inner side of the bottom part of the shell; the piezoelectric unit is fixedly connected with the matching piece; the lining plate is supported in the shell and isused for dividing the inner part of the shell into a sealing zone and a working zone; the matching piece and the piezoelectric unit are located under the lining plate and in the working zone; the potting adhesive is arranged in the sealing zone in a potting way and is located on the surface of the lining plate; the lining plate, the piezoelectric unit and the matching piece are fixedly arranged inthe shell by the potting adhesive in a sealing way; a signal line upward penetrates through the lining plate and the potting adhesive from the piezoelectric unit and stretches out of the shell; the lining plate and the piezoelectric unit are in clearance fit in a vertical direction of the inner part of the shell, and an air layer is formed between the lining plate and the piezoelectric unit. Thegas ultrasonic flow sensor disclosed by the invention is large in structural strength and good in sealing performance, and is capable of effectively ensuring fixing and sealing on the piezoelectric element and solving a shear force brought by expansion coefficients of all components; the original resolution ratio of a product cannot be reduced while the sensitivity is improved.

The invention provides a colour filter array and an image sensor. The colour filter array comprises multiple colour filter units; each colour filter unit is composed of total four rectangular pixels including a first pixel, a second pixel, a third pixel and a fourth pixel; the long side of each rectangular pixel is two times of the short side; the first pixel and the second pixel are sequentially appressed and arranged in a short side alignment manner along the first direction, where the long side is; the third pixel and the fourth pixel are sequentially appressed and arranged in a short side alignment manner along the first direction; and the first pixel and the third pixel are sequentially staggered, appressed and arranged by taking the length of the short side as the staggering distance along the second direction, where the short side is. By means of the colour filter array and the image sensor provided by the invention, on the premise that the resolution ratio of the image sensor is not reduced, the area of the unit colour filter pixel is increased to the most extent; and thus, the sensitivity of the image sensor is improved.

This invention relates to a color emendating method. The color measuring system respectively selects the plural groups of gray-rank value and the gray-rank representing value of the display lamp source of display, and respectively have numerical value calculation to the selected information of every color of light, then obtains the fitting curve function of gray-rank value data in every curve fitting section. Obtain the fitting gray-rank representing value of the gray-rank value in the sections using the fitting curve function, and tabular a comparison table. Besides, to correspond an unitary gamma curve of the image gray-rank formation to an object curve set beforehand. Logarithmize the two curves firstly when corresponding, get the emendated gray-rank signal through the comparison table. Output the emendated gray-rank signal to present the gray-rank distributing state by display. It's suitable for all kinds of display, especially for the liquid crystal display.

The application discloses a display device and a sub-pixel rendering method. The display device includes a plurality of pixel groups, the pixel group includes a first sub-pixel group and a second sub-pixel group; the first sub-pixel group includes a first sub-pixel, a second sub-pixel and a third sub-pixel; the second The sub-pixel group includes a first sub-pixel, a fourth sub-pixel and a fifth sub-pixel; wherein, the second sub-pixel is of the same color type as the fifth sub-pixel and has a different chromaticity; the third sub-pixel has the same color as the fourth sub-pixel Same type, different shades. By using sub-pixels of the same color type and different chromaticity in the sub-pixel arrangement, the transmittance of the display panel is improved without reducing the resolution and widening the color gamut.

A D / A converter in an analog-digital mixed loop is realized with a reduced circuit area without sacrificing the resolution. A bit modulation section modulates an m-bit digital control signal, which is output from a digital section, into an n-bit (n<m) intermediate digital signal whose temporal average precision is substantially m bits. A D / A conversion section converts the intermediate digital signal into an intermediate analog signal having a range corresponding to m bits. Then, by the smoothing operation through an analog filter, the intermediate analog signal is output as an analog control signal that has an m-bit precision as that of the original signal. In this way, it is possible to reduce the bit width for the internal operation and thus to reduce the circuit area without sacrificing the resolution of a D / A converter.