7results about How to "Improving Imaging Efficiency" patented technology

The application discloses a kind of acoustic-electric tomography image imaging algorithm based on two-point gradient iterative method, and the innovation is to improve the calculation efficiency by two-point gradient iterative method, and the accuracy of image reconstruction is improved by the selection of regularization term.The core of the application is to use the characteristics of acoustic-electric hybrid imaging, that is, when the boundary detection current is applied, the power density data is obtained by using ultrasonic disturbance, so as to obtain the internal data of biological tissue.At the same time, according to the different characteristics of different biological tissue conductivity, different regularization terms are introduced, two-point gradient iterative method is established, and high-resolution reconstruction of biological tissue conductivity is realized.Compared with the traditional acoustic-electric imaging method, the application reduces the calculation and storage cost of multiple observation currents, improves the imaging efficiency.And the iteration process is modified by using a special regularization term, so that the boundaries or geometric characteristics of different tissues can be accurately determined, and the imaging accuracy is improved.

This utility model relates to the field of medical device technology, and in particular to an intraoral scanning mounting structure and an intraoral scanning device. The intraoral scanning mounting structure includes: a housing with a mounting position, on which a protrusion is provided; an electro-optical component is mounted on the mounting position, and a groove is formed on the circumferential edge of the electro-optical component for the protrusion to extend into, thereby positioning the electro-optical component; an elastic component includes a first elastic member and a second elastic member, which are independently provided. The first elastic member is clamped between the mounting position and a first wall of the electro-optical component, and the second elastic member is clamped between the mounting position and a second wall of the electro-optical component, thus satisfying the installation and fixing requirements while allowing the electro-optical component to move within the mounting position due to thermal expansion and contraction; in the circumferential direction of the electro-optical component, a groove is provided between the portion of the first wall that abuts against the first elastic member and the portion of the second wall that abuts against the second elastic member, thereby ensuring the installation accuracy requirements of the electro-optical component.

The application relates to a deep learning microwave imaging method fusing virtual antenna data, comprising the following steps: a two-dimensional electromagnetic scattering model is established, and original scattering field observation data are collected under a sparse antenna configuration; the original scattering field observation data are processed by adopting a Fourier zero padding (FDZP) method, virtual scattering field data based on FDZP expansion are obtained, and a data set is further constructed; an FDZP-UNet network is constructed; the data set is input into the FDZP-UNet network, target dielectric constant prediction is output, and a high-resolution reconstructed image is further obtained. The deep learning microwave imaging method fusing virtual antenna data provided by the application comprehensively considers three performance indexes of imaging fidelity, system hardware cost and reconstruction efficiency, realizes virtual aperture expansion at a physical level through the Fourier domain zero padding method, effectively makes up for the physical information loss caused by sparse sampling, and greatly reduces the dependence on a large-scale physical antenna array.

This invention discloses a multi-channel light-slice microscopic imaging device. The image acquisition module includes multiple image detectors. A laser array emits multiple excitation beams of different wavelengths, which are then combined into a single incident beam by an optical path module. This incident beam passes through a scanning galvanometer and an imaging module before illuminating the sample on a stage mounted on a three-dimensional displacement platform. The imaging objective in the imaging module collects fluorescence signals from the sample, which are then split by the beam-splitting module to form multiple fluorescence signals of different wavelengths and projected onto the multiple image detectors. Each image detector forms an image imaging channel. Multiple image imaging channels receive multiple fluorescence signals of different wavelengths and transmit them to the corresponding image detectors. The three-dimensional displacement platform enables three-dimensional displacement movement of the sample slice. By employing a strategy of continuous motion imaging of the sample slice and simultaneous detection and acquisition of fluorescence signals by multiple image detectors, the time interval between sample preparation and imaging, as well as the pauses during imaging, can be reduced.

The application discloses a kind of single-pixel imaging method and device based on array light source, including obtaining the illumination light field of array light beam and the back light intensity detection value of the reflection back after array light beam irradiation on target object;Based on back light intensity detection value and illumination light field, reconstruct target image as rough target image;Based on rough target image and back light intensity detection value, reconstruct target image using depth neural network without training, output final target image.The present application realizes high efficiency imaging and long distance target detection based on array light source, while removing image noise, artifact and periodicity by introducing depth neural network, further improves the imaging quality and imaging efficiency of single-pixel imaging.

The invention discloses a CT image reconstruction method and system based on maximum residual error expansion oblique projection, and belongs to the technical field of medical image processing. Splitting the to-be-reconstructed CT image matrix according to columns to form a plurality of linear equations; setting the initial values of the auxiliary vector zk and the solution vector xk as z0 and x0; in the first iteration, a working column and a working row are selected according to a maximum residual index, and z0 and x0 are iterated through an orthogonal projection method; then working columns / rows continue to be selected according to the maximum residual error index, new column / row projection direction vectors are calculated through the working columns / rows of two adjacent steps, the auxiliary vector zk is iterated through the new column projection direction vectors and the appropriate step length, and if the new row projection direction vectors are not zero vectors, the new row projection direction vectors are not zero vectors; if so, iterating the solution vector xk through a new row projection direction vector and a proper step length, otherwise, iterating the solution vector xk through orthogonal projection; outputting a final solution vector xk + 1 when a termination condition is met; and combining all the final solution vectors according to the column xk + 1 to obtain a reconstructed CT image. The CT image reconstruction precision and the imaging efficiency are improved.

The application discloses a spectral super-resolution method based on a variational autoencoder, and comprises the following steps: inputting a target RGB image into a first convolutional layer of a spectral super-resolution network to obtain a plurality of feature maps containing shallow layer features corresponding to the target RGB image; inputting the feature maps into an attention mechanism module of the spectral super-resolution network to perform feature mapping on the feature maps; inputting the feature maps subjected to the feature mapping into a second convolutional layer of the spectral super-resolution network to reconstruct a target hyperspectral image corresponding to the target RGB image; and training the spectral super-resolution network according to a first hyperspectral image and a first RGB image extracted by the variational autoencoder, wherein the first hyperspectral image is reconstructed by the untrained spectral super-resolution network on the basis of a training RGB image. The target RGB image is acquired by a common camera, so that the image acquisition difficulty is low, and the hyperspectral image is reconstructed by the spectral super-resolution network, so that the imaging time is short, and the method can be widely applied to the field of spectral super-resolution.