50results about How to "Reduce human interaction" patented technology

The invention discloses an intelligent segmentation method integrating target detection and image segmentation. According to the method, a target is extracted by utilizing a high-dimension feature extracted by a neutral network; and a deep learning-based target detection network Faster-RCNN has high efficiency and can quickly and accurately identify a target region, so that a target region of interest can be extracted firstly by utilizing the characteristic and then the region is subjected to target segmentation pointedly. The method can automatically detect and extract the region of interest, reduce occupation of calculation resources of a GPU, and quickly and accurately perform segmentation to obtain the target; and under the condition of a huge image quantity scale, the method has efficient implementation performance, so that the manual interaction process can be reduced.

The present invention relates to an automatic segmentation method for a breast MRI focus based on inter-frame correlation. The method comprises: reading an MRI image; preprocessing the image, and performing coarse segmentation on the preprocessed image to determine an initial contour of a focus; and by adopting an improved C- V level set model method, performing fine segmentation on a coarsely segmented image that is obtained in the prior step, further refining the tumor contour on the basis of a coarsely segmented contour, and optimizing an obtained finely segmented result by combining inter-frame correlation. The method provided by the present invention has higher accuracy and can better segment the focus.

The invention discloses a blue screen image-matting method. The blue screen image-matting method comprises the following steps of: transferring a video into a GPU, and carrying out blue screen image-matting processing on a video frame in the GPU, thus increasing the image-matting speed. Specifically, the blue screen image-matting method comprises the following steps of: extracting a background colour from the video frame, carrying out non-transparency processing on a blue screen image by virtue of a chromatic aberration image-matting technology, then obtaining an initial non-transparency image through binarization processing, then optimizing the initial non-transparency by virtue of the gradient information of a colour image and an improved combined bilateral filter to obtain a final non-transparency image, and finally carrying out background colour overflow removal on the final non-transparency image and then returning the final non-transparency image to a CPU client, and displaying through a display device. In this way, human interaction and parameter adjustment are greatly reduced during the whole processing process.

The present invention discloses a video editing method based on concatenation of panoramagram, being used for editing or repairing section series of motion video. The method includes the steps as follows: matching with the characteristic point of the every frame image of the video series, calculating the projection matrix between the frames and appending the image to acquire the video panoragram; editing the video panoragram with the human interaction and computer according to the specific demand of video editing; restoring the video series from the edited panoragram based on the projection relationship. Applying the video panoragram as the medium, the present invention shifts the video editing into the image editing of vider panoragram, not only directly providing the whole scene message to the user, but also reducing the calculating quantity and the human interaction.

The invention discloses a method for translating an SAR image into an optical image. The method includes the following steps that: a translation knowledge base is built; preprocessing and feature extraction are performed on an SAR image to be translated; characteristics and a classification algorithm are utilized to carry out image classification, the types of surface features are marked; the mapping relationship between the SAR image to be translated and a target image surface feature is determined according to a classification result; a target expression method is adopted to perform conversion and re-expression of a surface feature target on the SAR image to be translated according to a translation rule and a target sample provided by the translation knowledge base; and an optical image obtained through translation is outputted. The method of the invention has the advantages of reasonable process, high efficiency of the algorithm, little man-machine interaction and excellent translation effect. With the method of the invention adopted, a translation result image is as consistent as possible with an SAR image in the aspect of maintaining surface feature spatial structure information, and the translation result image is very close to a target image in aspects such as spectrum, hue and texture features.

The invention discloses an image enhancing method based on self-adaptive block channel stretching. The method mainly solves two important problems of image defogging and exposure adjustment in image enhancement. In the prior art, many methods provide solution schemes for the two problems, but the methods have the defect that a plurality of thresholds for image adjustment are needed, and this kind of methods requires user interaction operation and fail to process a plurality of images at one time; in addition, in the prior art, fogged images and dark light images are taken as two kinds of image enhancing problems which are solved respectively, and a unified framework is lacked. Aimed at the defects of the existing methods, statistic learning is carried out on various kinds of images in a real world, a self-adaptive threshold of problem image adjustment is obtained, the manual interaction operation of a user is reduced, and the processing efficiency is improved; in addition, stretching conversion is simultaneously carried out on a highlight channel and a shadow channel of an image block, and the problem images including the fogged images and the dark light images can be enhanced under the same framework.

The invention discloses an iteration generation method of a hexahedral mesh model. The iteration generation method comprises the following steps: step one, inputting a triangular mesh model, constructing an initial hexahedral mesh model according to the triangular mesh model, and extracting a surface quadrilateral mesh from the initial hexahedral mesh model; step 2, carrying out iterative fitting of the surface quadrilateral mesh of the initial hexahedral mesh model on the triangular mesh model and carrying out mobile diffusion on mesh peaks of the surface quadrilateral mesh so as to obtain an intermediate hexahedral mesh model that is divided into multiple layers of quadrilateral meshes; and step three, inserting a layer of quadrilateral mesh between the surface quadrilateral mesh of the intermediate hexahedral mesh model and the internal first-layer quadrilateral mesh and optimizing all layers of the quadrilateral meshes except the surface quadrilateral mesh so as to obtain a final hexahedral mesh model. According to the invention, with the method, a hexahedral mesh with the guaranteed quality can be obtained; and the automation degree is improved.

The invention relates to a blue screen image matting method based on chroma overflowing processing. The blue screen image matting method is characterized by comprising the following specific steps: dividing a foreground image to be subjected image matting into a non-transparent region, a semi-transparent region and a full-transparent region, and displaying a region division result; calculating based on a sampled Alpha value; calculating the Alpha value of each point of the foreground image; loading a background image used for being synthesized with the foreground image; synthesizing the foreground image and the background image; calculating a RGB color value of each pixel point of the synthesized image; and carrying out the chroma overflowing processing on each pixel point of the synthesized image. According to the method, only simple input needs to be carried out before photographing, namely the image matting and the chroma overflowing processing can be automatically finished; with regard to the image with the more serious chroma overflowing phenomenon, the more satisfied effect can be obtained.

A medication dispenser provides automation to the steps of locating and acquiring unit-based doses of certain medications to be administered to a patient. The dispenser includes a frame and one or more cartridges that may be mounted onto the frame. A set of slots sized for holding unit-based doses of medication extend through a body portion of the cartridge. A movement device is also positioned relative to the frame and is configured to induce movement of selected unit-based medication doses out of associated slots in the cartridge, so that the dispensed doses may be retrieved.

The invention relates to an arteriovenous malformation segmentation method and system and electronic equipment. The arteriovenous malformation segmentation method comprises the following steps: a, segmenting a blood vessel tree from a nuclear magnetic resonance angiography image to generate a blood vessel skeleton line; b, creating a spanning tree of a vascular skeleton line by using a weight-based breadth-first search algorithm, and constructing a topological path from a carotid artery root point to an AVM focus through a blood supply artery and then to a drainage vein; and c, positioning andsegmenting the AVM focus according to the spanning tree, and segmenting the blood supply artery and the drainage vein according to the topological paths of the blood supply artery and the drainage vein connected with the AVM focus. Compared with the prior art, the arteriovenous malformation segmentation method has the advantages that less human interaction is needed, and arteriovenous malformation can be automatically segmented, and AVMs with gaps and even sparse blood vessels can be well segmented, and blood supply arteries and drainage veins can be accurately segmented according to the topological relation of the blood vessels.

A single-tooth extraction method based on a panorama is characterized by comprising the following steps: step 1, extracting an alveolar bone foreground image; step 2, generating a tooth panorama according to the alveolar bone foreground image obtained in the step 1; step 3, generating an upper and lower jaw separation line according to the tooth panorama obtained in the step 2; step 4, generatingadjacent tooth segmentation lines according to the tooth panorama obtained in the step 2; step 5, extracting a single tooth from the three-dimensional oral CT by using the upper and lower jaw separation line obtained in the step 3 and the adjacent tooth segmentation line obtained in the step 4; and step 6, carrying out post-processing on the generated three-dimensional single-tooth CT image stack.The method for extracting the single tooth based on the panorama is easy to implement and high in robustness, is not limited to a CT image of the oral cavity of a patient with the open oral cavity, and can extract any single tooth in the oral cavity of the patient. And the method realizes automatic segmentation of adjacent teeth, and manual interaction is reduced as much as possible.

The invention provides a DOM local repair polymerization method with balanced and non-deformable colors. The influence of the boundary on the aggregated image is fully considered; in order to ensure that the foreground color is distortionless, the foreground and the background are optimally sewed, and the background color and structure are kept relatively complete. Three boundaries including a foreground fixed boundary, an optimal splicing line and a background fixed boundary are sequentially extracted; the method is characterized in that three boundaries are taken as contour lines, discrete points are constructed into an irregular triangulation network with constraints, and then weight value interpolation is carried out on the irregular triangulation network for obtaining a smooth weightvalue in the transition connection area; the aggregation result fully shows that the method can effectively improve the defects of color bleeding effect, color distortion and the like of the DOM, smooth transition is achieved, and deformation and color imbalance areas in the image can be effectively repaired.

The invention discloses a method for segmenting an aortic-valve ultrasound image sequence based on an interframe-shape-constraint GCV model. The method includes the following steps: acquiring a group of ultrasound images and carrying out Wiener filtering; calculating a gradient vector flow field of each image frame by frame and adding the gradient vector flow fields to a CV model as energy constraints and obtaining the GCV model; through defining an initial constraint shape and adding the initial constraint shape into the GCV model as an energy constraint item and then minimizing an energy functional so that a segmentation result of a first frame of image is obtained; and carrying out filtering of a rolling sphere method on the aortic-valve segmentation result of an adjacent previous-frame image and adding the obtained result into the GCV model as an energy constraint item so that the segmentation result of a current frame is obtained through calculation. The method for segmenting the aortic-valve ultrasound image sequence based on the interframe-shape-constraint GCV model is operated on short-axis images of ultrasonic cardiograms so that not only is work load of doctors reduced, a problem that serious overflow in segmentation of the aortic-valve ultrasound images in the prior art is also solved. The segmentation result is extremely close to the result of manual segmentation. The method enables the aortic valve to be segmented in a simple and highly-efficient manner.

An advisory system of a vessel that monitors variables of a vessel system inclusive of systems and subsystems that are used to operate the vessel. The advisory system may use machine-learning to learn from an operator (i) whether or not two variables are related to one another, and (ii) likelihood that a variable will reach a threshold, and, optionally, time until reaching the threshold. The system may receive operator feedback (i) to indicate whether the two variables are related to one another, and (ii) whether a behavior of the variable is normal or not normal. Thereafter, if a determination that the same two variables are related to one another and behaving in a similar manner, provide notification to the operator of the behavior. In response to determining that the variable is behaving (e.g., trending) in a similar manner that is not normal, providing a notification to the operator.

The invention discloses a rapid garment fabric filling method based on an edge matrix. The method includes the steps: storing garment area information and garment edge information; filling garment fabrics; treating a fabric filling area edge. According to the requirements of two-dimensional garment virtual display, man-machine interaction in garment fabric filling is decreased, the filling process of the garment fabrics is optimized, and the method solves the problem that the garment fabric filling is difficult and time is wasted in the garment virtual display to some extent.

The invention discloses a method for automatically and remotely executing a Windows program, which comprises the following steps of: mapping a local file directory to a server through a Windows remotedesktop protocol channel, and executing the Windows program or a Windows built-in program in the mapped file directory; and automatically and periodically logging in the Windows server in batches through scripts to execute the target program, and outputting a program execution result. According to the method, a script is created for each task and is used for logging in different Windows servers to execute the tasks in the scripts; and if the task needs to be altered, the change of task requirements can be quickly adapted only by modifying script content, the operation and maintenance operation is automatically executed, the manual interaction is reduced, the operation and maintenance work efficiency is improved, the operation and maintenance effect of the ideal Linux server is completed,the Windows server is logged in automatically and periodically through scripts in batches to execute target programs, and the purpose of operation and maintenance is achieved.

The invention relates to the technical field of image processing, and discloses a calibration point identification method, which comprises the steps of processing image data, and generating an external rectangle of a calibration object in an image, wherein the lower edge of the circumscribed rectangle extends downwards by a preset pixel coordinate length to generate a region of interest; traversing the region of interest to generate a marking region of a preset marker related to the calibration object; traversing the marking area to generate a middle line, and taking the pixel coordinate of the upper end point of the middle line as the pixel coordinate of the calibration point. The method has the advantages that manual interaction is reduced, and effective and automatic identification of the calibration points is achieved.

A tubular structure rapid tracking method based on curvature regularization perception grouping is disclosed. According to the method, prior information of local smooth change of tubular structure features is introduced into a model, and curvature regularization geodesic line model beam area curvature change is used for solving the problem of short cutting existing in tubular structure center line tracking of a geodesic line model; and a traditional geodesic line model-based pixel-by-pixel tracking mode is expanded into a curvature constraint-based pre-segmentation center line segment grouping tracking mode, so that the human interaction is effectively reduced, and the tracking speed and the accuracy in complex structure tracking are greatly improved.

In a desktop for workflow management, a global task list of tasks to be performed is generated. The global task list includes a sequence of plural individual task lists ordered according to a configurable task list order. Each individual task list contains one or more tasks ordered according to respective individual task list sort criteria, and each individual task list is updated automatically through a notification mechanism. For each individual task list in the global task list, respective tasks that form part of the individual task list are listed in the global task list in order of appearance in the individual task list.

The invention discloses a BMC system new user creation function test method. The method comprises the steps of obtaining to-be-created user information; executing a corresponding user automatic creation instruction; obtaining creation feedback information after the creation instruction is completed, performing creation function judgment according to the creation feedback information, wherein the test result creation comprises creation success and creation failure; judging whether the currently created user succeeds or not according to the creation test result. According to the method, a largeamount of manpower, time, cost and other resources are saved. The unnecessary manual interaction processes are reduced. The test efficiency is greatly improved. The correctness of the test operation is ensured. The method has good popularization and use values. The invention further provides a device and equipment for testing the new user creation function of the BMC system and a readable storagemedium, which have the above beneficial effects.

The invention relates to a method for synchronously segmenting multiple organs of a lung. The method comprises the following steps: a, performing interactive data annotation on the multiple organs of the lung; b, performing data preprocessing on the data annotation; c, constructing a data set for the data preprocessing result; d, enhancing and expanding the data set; e, carrying out DC-U-Net network model training on the data set to obtain a DC-U-Net training model; f, carrying out automatic segmentation on multiple organs of the lung based on the DC-U-Net training model. According to the invention, lung parenchyma, blood vessel, bronchus and pulmonary nodule areas can be automatically extracted based on the deep learning technology, and synchronous segmentation is realized.