13 results about "Image simplification" patented technology

Systems and methods are provided for medical image classification of images from varying sources. A set of microscopic medical images are acquired, and a first neural network module configured to reduce each of the set of microscopic medical images to a feature representation is generated. The first neural network module, a second neural network module, and a third neural network module are trained on at least a subset of the set of microscopic medical images. The second neural network module is trained to receive feature representation associated with an image of the microscopic images and classify the image into one of a first plurality of output classes. The third neural network module is trained to receive the feature representation, classify the image into one of a second plurality of output classes based on the feature representation, and provide feedback to the first neural network module.

The invention discloses a note illustration method and device, equipment and a storage medium, and belongs to the technical field of software auxiliary application. The method comprises the steps of displaying a note interface, and determining at least one handwriting of a to-be-illustrated picture based on the note interface; acquiring an illustration image matched with the at least one handwriting; and displaying the illustration image on the note interface. According to the method, the illustration image can be automatically obtained and displayed through the at least one handwriting of the to-be-illustrated image, the note illustration operation process is simplified, and the note illustration efficiency is improved. Moreover, the automatically acquired illustration image is matched with at least one handwriting, so that the accuracy of the illustration image is relatively high.

This invention provides an imaging analysis system, comprising: a housing; a sample stage mounted on the housing, the upper surface of the sample stage having a mounting portion for placing a culture unit, and the sample stage also having a vertically penetrating channel located within the area of ​​the mounting portion; and an imaging module located below the sample stage and connected to the lower surface of the sample stage, the imaging module being used to photograph the culture unit through the channel. By placing the mounting portion for placing the culture unit on the upper surface of the sample stage and connecting the imaging module to the lower surface of the sample stage, the imaging module can photograph the culture unit from below through the channel on the sample stage, thus obtaining a clear image of the culture unit. This simplifies the system structure, reduces the system size, and allows the culture unit and imaging module to be mounted on the same component (sample stage), reducing the machining and assembly precision requirements of the components, while also ensuring clear imaging by the imaging module.

The invention relates to an image display method, device and equipment, a storage medium and a chip system, and relates to the technical field of image display, the method can be applied to an intelligent cabin scene, and the method comprises the following steps: in response to receiving a first to-be-displayed image parameter sent by a microprocessing unit, rendering the to-be-displayed image parameter to obtain a rendered first image, and sending the first image to the display module, so that the display module processes the first image to obtain a second image, obtains a second to-be-displayed image parameter and the second image sent by the microprocessing unit, and displays the second image in response to the matching of the second image and the second to-be-displayed image parameter, thereby simplifying the processing flow and reducing the image display cost.

The invention discloses an image processing method and device and electronic equipment, and belongs to the technical field of image processing. The method comprises the steps that the position of a target fixation point is determined based on the pupil center position and the cornea reflection point position of a first frame image and the pupil center position of at least one second frame image, and the at least one second frame image is at least one frame image before the first frame image; taking the target fixation point position as a circle center, delimiting at least two concentric areas with different radiuses in the first frame image, the at least two concentric areas including a fixation area and a non-fixation area at the periphery of the fixation area; and performing image enhancement processing on the gaze area, performing image simplification processing on the non-gaze area, and outputting an image.

The invention discloses an image cutting method, image acquisition equipment and a storage medium, and relates to the technical field of image processing, and the image cutting method comprises the steps that a face detection head based on a CNN determines a face existence detection result of a current image shot by the image acquisition equipment; determining predicted cutting point coordinates according to a cutting point positioning head parallel to the face detection head in the CNN; and if the human face existence detection result is that the human face exists, cutting the current image based on the predicted cutting point coordinates. After the image is shot, face existence detection and cutting coordinate positioning are synchronously completed through the architecture that double task heads of the CNN work in parallel, the image can be automatically cut without user operation, the image does not need to be screened one by one, the image processing flow is simplified, and the image processing efficiency is improved.

The invention relates to the technical field of image processing and transmission, and discloses an edge-end-oriented image simplification algorithm, which comprises the following modules: an edge-end data acquisition module, a server relay module and a client presentation module. Two channels of a simplified frame and an original frame are established through the edge end data acquisition module, a double-queue buffer system is established through the server relay module, two-channel data of the simplified frame and the original frame are received, and the client presentation module processes received frame pairs of the server relay module. And sending a data request for the specific edge end data acquisition module to the server relay module for interaction. Through cooperation of video stream frame classification, key visual information feature extraction and compression transmission and in combination with an interactive region request, key visual information is reserved while data optimization transmission is realized.

The application provides a face image generation method and device, electronic equipment and storage medium. The method comprises the following steps: obtaining first classification features of an initial face image by a first sub-classification model of a first neural network obtained through training; determining parameters of a first sub-flow model of the first neural network based on the first classification features, and inputting the target face image into the first sub-flow model with the determined parameters for forward operation to obtain initial features of the initial face image; obtaining first condition information input by a user, adjusting the initial features based on the first condition information; inputting the adjusted initial features into the first sub-flow model with the determined parameters for inverse operation to obtain a first target face image corresponding to the adjusted initial features, so that the target face image can be quickly obtained according to the initial face image, the user's face adjusting operation is simplified, and the user can quickly obtain the attribute features that he or she wants.

The present application relates to a phase retrieval wavefront detection method and system, in particular to a phase retrieval wavefront detection method and system based on a non-standard symmetric pupil, which solves the technical problem of poor precision and stability of the existing phase retrieval wavefront detection method using a standard symmetric circular pupil. The present application uses a non-standard symmetric pupil, so that the optical transfer function presents a distribution with obvious direction characteristics in the frequency domain, thereby enhancing the direction sensitivity of the imaging result to system aberration and misalignment parameters, and more abundant aberration information can be extracted from the imaging data to provide stronger constraint conditions for wavefront recovery, and high-precision and high-stability wavefront detection is realized under complex imaging environment. At the same time, imaging information with direction characteristics can be introduced into the imaging data, thereby realizing effective detection of the wavefront, so that images in different states do not need to be obtained by mechanically moving the focal plane, and the system structure is simplified.

The present invention relates to the technical fields of life sciences and medical image devices, and particularly relates to a method for generating the internal structure of an organism on the basis of deep learning. The method comprises the following steps: S1, constructing and training an internal organ generation model of an imaging target; S2, acquiring a three-dimensional body surface contour image of the imaging target; S3, inputting the acquired three-dimensional body surface contour image of the imaging target into the internal organ generation model, so as to obtain a three-dimensional internal organ distribution image; and S4, fusing and displaying the three-dimensional internal organ distribution image and the three-dimensional body surface contour image. By means of the deep-neural-network-based internal organ generation model of the imaging target, it is only necessary to acquire the three-dimensional body surface contour image of the imaging target, and the three-dimensional internal organ distribution image of the imaging target can be quickly and accurately acquired after the three-dimensional body surface contour image is input into the model, thereby simplifying operation steps, realizing automatic imaging and improving the imaging efficiency. Therefore, the method has a strong generalization performance.

A robot control method for controlling a robot on the basis of simplified maps and natural language commands according to an embodiment of the present invention may comprise the steps of: acquiring an image regarding a space; acquiring a simplified map of the space including the robot's current location and location to be reached, at least some elements depicting the space being omitted from the simplified map; acquiring a command expressed in a natural language, which is to be observed by the robot during operations; generating an input token on the basis of at least one of the image, the simplified map, and the command; and generating a waypoint token by processing the input token with a trained vision-language-action model.

The invention provides an image processing method and device and a microscopic camera. The target area is determined from the reference image of the target object collected by the image collection device based on the area selection instruction, the translation stage is controlled to move so that the target area can be aligned with the visual field of the image collection device, related personnel do not need to manually adjust the position of the target area, the positioning efficiency of the target area is improved, and the user experience is improved. And errors caused by manual operation are eliminated. Furthermore, by controlling the translation stage to move along the optical axis direction of the image acquisition device, the target position of the translation stage in the optical axis direction of the image acquisition device when the target area is in the focusing state is determined, and the translation stage is controlled to perform automatic scanning based on the target position so as to generate a panoramic depth image of the target area. The panoramic depth image can be automatically obtained only through one-time operation, the operation complexity is simplified, and the generation efficiency of the panoramic depth image is improved.

The invention provides a BIM model standard layer identification method and system based on plane projection comparison, a program product and a storage medium, and relates to the technical field of BIM model identification. Cutting the BIM model according to floor boundaries to obtain a plurality of floor models, then dividing the floor models with the same floor height into one group, performing vertical projection under the same coordinate system for each floor model group, converting a three-dimensional floor model into a two-dimensional floor projection image, and performing three-dimensional projection on the two-dimensional floor projection image; the data form is simplified while preserving information based on subsequent pixel comparison. Cutting the image according to a preset grid to generate a plurality of image units, comparing the image units at the same grid position in the same floor model group in pairs to judge a standard floor, and efficiently and accurately completing standard floor identification. The problems of misjudgment and workload increase in application caused by lack of standard layer information when the BIM model is converted or imported in the past are solved, and the overall working efficiency and the standard layer identification accuracy are improved.