32results about How to "Automatic segmentation" patented technology

The invention discloses an ultrasonic image segmentation method of cascaded total convolution neural network for thyroid nodule, which comprises the following steps of: constructing an ultrasonic image segmentation method based on U-NET, the region of interest (ROI) is segmented from the ultrasound images of thyroid gland based on the simplified total convolution neural network (TCNN); using the VGG19-FCN network as the down-sampling layer to extract the deep features of ROI, so as to realize the automatic semantic segmentation of thyroid nodules. The simplified total convolution neural network comprises a fifth convolution layer for downsampling and a fifth upsampling layer for upsampling. The first five convolutions conv are composed of two 3x3 convolution layers and one pooling layer, each convolution layer uses ReLU as the activation function, and the last five upsampling layers are deconvolution layers. The invention provides a high-precision nodule image for identifying benign and malignant thyroid nodules, thereby playing a better auxiliary role in medical diagnosis.

The invention provides a division method and device of left ventricular myocardium. The method comprises the following steps of: (1) positioning a left ventricular central point, and acquiring a region of interest in an image at a layer where the left ventricular central point is located; (2) positioning a blood pool region in the region of interest in the image at the layer where the left ventricular central point is located according to grey scale information and the left ventricular central point; (3) determining a mass center of the blood pool region, determining blood pool regions in an original image in all layers except the layer where the left ventricular central point is located, thereby obtaining a division result of left ventricular endocardium; and (4) obtaining a division result of left ventricular adventitia by a dynamic planning method according to the division result of the left ventricular endocardium, and thus obtaining a division result of the left ventricular myocardium. By the technical scheme provided by the invention, the left ventricular myocardium can be fully, uniformly and stably divided.

The present invention discloses an airborne Lidar point cloud building top surface gradual extraction method based on classifying and laying. The method comprises the steps: firstly classifying top surfaces of a building; according to the size of top surface area, dividing the top surface into a "big top surface" and a "small top surface", and according to the size of an angle between the top surface, dividing the top surfaces into different levels. On this basis, the method adopts the principle of "from big to small", "from rough to fine" and "classified process", gradual extracting the top surface of building from LiDAR point cloud. The method comprises: firstly, combining a region growing method based on normal and a region growing method based on distance, segmenting the big top surface from point cloud; then performing clustering to the remaining points, and segmenting the small top surface from each cluster by using a random sample consensus method; and by constantly improving the determination condition of angles between the top surfaces, and gradually segmenting the finer angle between the top surface. and finally, achieving automatic and accurate segmentation for various top surfaces of buildings, thereby laying a foundation for automatic modeling of three-dimensional buildings.

The invention relates to the technical field of pelvis and pelvis artery blood vessel tree segmentation, in particular to a deep learning-based pelvis and pelvis artery blood vessel multi-level segmentation method, which is used for solving the problem that an abdomen pelvis and a pelvis artery blood vessel tree cannot be automatically, efficiently and accurately segmented in a multi-resolution CTimage in the prior art. The method comprises the following steps: step 1, preparing and labeling data; 2, preprocessing the data; 3, constructing a first-level segmentation model of the 3D convolutional neural network based on multi-level segmentation; 4, constructing a second-stage segmentation model; 5, training a first-stage segmentation model and a second-stage segmentation model by using thecalibrated data and the synthesized loss function; and 6, performing abdomen information segmentation on the input three-dimensional CT image by using the first-stage segmentation model and the second-stage segmentation model trained in the step 5. According to the method, the abdominal pelvis and the pelvis vascular tree can be automatically, efficiently and accurately segmented in the multi-resolution CT image.

The invention belongs to the technical field of image segmentation, and specifically relates to a 3D full-automatic glioma segmentation method combining a Volume of Interest and a GrowCut algorithm. The method first expands a Bounding box algorithm to 3D, utilizes the algorithm to extract a Volume of Interest (VOI) containing glioma, then utilizes a reflection symmetry algorithm to estimate the VOI and overcomes the defects of Bounding box when the glioma is detected to cross over a median sagittal plane, and finally marks pixel points in an image based on the accurate VOI, and thus a semi-automatic 2D GrowCut algorithm is optimized to a full-automatic 3D segmentation method. While accurately segmenting the glioma, the method is more rapid in theory and reality compared with the 2D segmentation algorithm of the same principle, and is better in convenience and feasibility compared with a manual segmentation method. As an image segmentation method, the 3D full-automatic glioma segmentation method provided by the invention can serve as a powerful auxiliary tool for clinical diagnosis of glioma.

The invention relates to a dual-source CT coronary artery automatic extraction method. The method comprises the following steps of S1, calculating eigenvalues lambda 1, lambda 2 and lambda 3 of a Hessian matrix for each pixel point in dual-source CT data, and building a new blood vessel model equation; effectively enhancing a blood vessel structure by utilizing the new blood vessel model; enablinga blood vessel region to be relatively bright in grayscale, and enabling a background region to be relatively dark; performing Gaussian modeling on the blood vessel region and the background region by using a statistical model; S2, by utilizing a Gaussian mixture model, classifying enhanced data into two types, namely, the background region and the blood vessel region; and S3, in the blood vesselregion, restraining a neighborhood relation through the growth of a ball model, growing coronary arteries, and detecting and removing pseudo-branch blood vessels in the grown coronary arteries by utilizing hierarchical clustering-based branch detection.

The invention discloses a splitting method of a CD image of a solid engine. The method comprises the following steps: a CD faultage image data of the solid engine is read in, an artifact which is formed outside the solid engine owning to the air agitation formed in the CD detection is eliminated; the image, the artifact of which is eliminated, filters and eliminates background noise by median filter, meanwhile, the edges of each component of the solid engine are sharpened in the CD image and a casing, a propelling agent and an astropyle of the solid engine is cut in the CD image; a defect is cut from the propelling agent. By adopting the method, an outer layer air ring artifact, the casing, the propelling agent, the astropyle and various defects of the solid engine in the CD image are cut automatically and accurately, thus solving the problem of low efficiency and poor accuracy in the CD detection image of the solid engine by the manual interpretation as well as being not limited by the eigenvalue of image information.

The invention discloses an image processing method applied to surgical navigation. The method comprises the steps that a tested image is acquired; if the acquired tested image comprises a multimodality image of a CT image and / or an MR image, multimodality image registration is performed; voxels of the tested image are segmented and marked; features of background, vessels, nerves, skeletons, foci and visceral organs where the foci are located are extracted, and feature vector preprocessing is performed; feature parameters of focus areas obtained after image segmentation are calculated; the end point of a surgical navigation path is determined though feature parameter calculation, a starting point area of the surgical navigation path is determined in a man-machine interaction mode, and path optimization is done. Through multimodality image registration and analysis, automatic and accurate segmentation of the foci, calculation of the parameters of the foci and output of an optimized path for cooperation with visual operation during surgical navigation are realized.

The invention relates to an echocardiographic ventricle segmentation method and device based on deep learning and a deformation model, and aims at ironing out a defect that a conventional manual boundary marking method exerts a big negative impact on the calculation of the related indexes of a ventricle because the conventional manual boundary marking method usually consumes a large amount of manpower and material resources and the marking results of different persons are different. The method comprises the steps: carrying out the training of a ventricle coarse segmentation model through manually marked training data, and obtaining a coarse segmentation training result; calculating the central point of the coarse segmentation training result on each section, carrying out the fitting of oneline through all central points, and calculating the mean value of the distances between the central points in a direction perpendicular to the line and an outer edge of the coarse segmentation training result as the radius; carrying out the resampling according to the calculated central points and the radius, and reconstructing a three-dimensional initialization model based on a sampling result;and carrying out the fine segmentation of the ventricle coarse segmentation result through the deformation model. The method is suitable for the ventricle image processing.

An image processing method for converting 2D (2-dimensional) video into 3D (3-dimensional) video comprises the steps of taking images, shot from different angles, of a target object as sample images, calculating color invariable images of the sample images and a to-be-identified image, extracting feature points from the color invariable images, matching the feature points in the color invariable image of the to-be-identified image with the feature points in the color invariable images of the sample images, judging whether the target object exists in the to-be-identified image according to distribution of the feature points matched successfully, if the target object exists in the to-be-identified image, dividing a target object outline based on the feature points matched successfully. The method can identify the target object when the 2D video is converted into the 3D video, precisely divides the object outline, and greatly reduces the workload of manual operation.

The invention relates to a pulmonary nodule segmentation method for a lung CT image, and the method comprises the following steps: obtaining a lung CT image data set, and carrying out the data enhancement of the lung CT image data set; dividing the lung CT image data set after the data enhancement operation into a training set and a test set; constructing a semantic segmentation model based on the enhanced UNet structure, and training and testing the semantic segmentation model by using the training set and the test set to obtain a pulmonary nodule segmentation model; and inputting the actual lung CT image to be measured into the pulmonary nodule segmentation model, and outputting to obtain a corresponding pulmonary nodule segmentation image. Compared with the prior art, the pulmonary nodule segmentation method has the advantages that deep learning and a semantic segmentation technology are combined, an end-to-end semantic segmentation model is provided, and the lung and pulmonary nodule can be automatically, quickly and accurately segmented from the lung CT image by utilizing the ResNet101 backbone network, the context information extraction module and the spatial information extraction module in the semantic segmentation model.

An integral graph algorithm-based fabric flaw detection method is disclosed. An integral graph algorithm is used for rapidly extracting statistical characteristics of gradient energy, and the statistical characteristics of gradient energy are used for flaw detection; image learning operation on a flawless template is performed, statistics are run on characteristic distribution of the gradient energy, distribution peak values are extracted, threshold parameters are calculated in a self-adaptive manner and used for distinguishing subsequent flaws, gradient energy of a window where each pixel point of an image to be detected is positioned can be calculated via the integral graph algorithm, whether a current pixel point is a defect point can be determined based on the threshold parameters, and whether the current image is a flawed fabric can be determined after statistics are run on a total quantity of flaw points of the whole image. According to the method disclosed in the invention, based on principles of accelerated operation of integral graphs, the characteristic distribution of the gradient energy of the fabric image can be rapidly extracted, real time detection of fabric flaws can be realized, the peak values of the distribution are calculated, the self-adaptive threshold parameters for flaw determination are obtained, and accurate segmentation of fabric flaws can be realized. Via the method, real-time property and high accuracy can be ensured.

The invention relates to a probability graphic model image segmentation method based on flag fusion. The method comprises the steps of (1) acquiring a target image and multiple training images, conducting registration on the target image and the training images, and obtaining a corresponding deformation field; (2) acquiring the manual segmentation flag image of each training image subjected to registration, and obtaining a corresponding candidate flag according to deformation field reflection; (3) establishing a probability graphic model according to the training images and target image subjected to registration; (4) calculating the flag value of each pixel point in the target image by means of the local weighted voting flag fusion algorithm and the maximum posterior probability according to the probability graphic model, so that segmentation of the target image is achieved. Compared with the prior art, the method has the advantages that segmentation can be achieved automatically and segmentation precision is high.

The invention discloses a sectional water feeding type solar water heater. A water storage tank (1) is connected with two water outlets, one water outlet is positioned in the top end of the water storage tank and connected with a water return pipe (4) through a pipeline, and a water outlet valve SV1 is arranged on the pipeline of the water outlet; the other water outlet is positioned in the middleof the water storage tank and connected with the water return pipe through a pipeline, and a water outlet valve SV2 is arranged on the pipeline of the water outlet; water of the water return pipe flows to a water bucket (2), and the top end of the water bucket is connected with a movable arm (5) used for controlling a travel switch SQ to be switched on and switched off; and the solar water heaterfurther comprises a thermistor Rt1, a photoresistor Rs1 and a peripheral circuit, wherein the thermistor Rt1 is mounted in the water storage tank and used for sensing the water temperature, the photoresistor Rs1 is used for sensing the illumination intensity, and the peripheral circuit is connected with the thermistor Rt1, the photoresistor Rs1, the water outlet valve SV1, the water outlet valveSV2, a water inlet valve SV3 and the travel switch SQ. According to the solar water heater, the water feeding level can be automatically selected according to the illumination intensity of sunlight and the water temperature condition.

The present invention relates to the technical field of segmentation of pelvis and pelvic artery tree, in particular to a multi-level segmentation method of pelvis and its arterial vessels based on deep learning, which is used to solve the problem of multi-resolution CT in the prior art. The problem of automatic, efficient, and accurate segmentation of the abdominal pelvis and pelvic artery vascular tree in images. The present invention includes step 1: data preparation and labeling; step 2: data preprocessing; step 3: construction of a first-level segmentation model based on a multi-level segmentation 3D convolutional neural network; step 4: construction of a second-level segmentation model; 5: Use the calibrated data and the synthetic loss function to train the first-level segmentation model and the second-level segmentation model; Step 6: Use the first-level segmentation model and the second-level segmentation model trained in step 5 to input 3D CT images Perform abdominal information segmentation. In the present invention, the abdominal pelvis and the pelvic vascular tree can be automatically, efficiently and accurately segmented in multi-resolution CT images.

The invention belongs to the field of computer-aided diagnosis, medical image processing, ultrasonic tomography and medical image analysis, and particularly discloses an anti-artifact feature extraction method based on an ultrasonic tomography reflection image convex target, which comprises the following steps of: (1) drawing an initial convex hull aiming at the convex target in a mode of selecting nodes; (2) calculating a geometric gravity center O1 of a target representation pixel; (3) updating and iterating the nodes; and (4) fitting and shaping the nodes, wherein a region correspondingly surrounded by a contour line obtained by connecting the shaped new nodes is a corresponding image of the extracted convex target. According to the method, details such as the overall process design and iteration conditions of the method are improved, so that the anti-artifact feature extraction method based on the ultrasonic tomography reflection image convex target can be accurate and effective in feature extraction for the reflection image of medical ultrasonic tomography equipment, and the extraction efficiency is improved. And meanwhile, the method has good robustness for artifacts and noise.

The present invention discloses an airborne Lidar point cloud building top surface gradual extraction method based on classifying and laying. The method comprises the steps: firstly classifying top surfaces of a building; according to the size of top surface area, dividing the top surface into a "big top surface" and a "small top surface", and according to the size of an angle between the top surface, dividing the top surfaces into different levels. On this basis, the method adopts the principle of "from big to small", "from rough to fine" and "classified process", gradual extracting the top surface of building from LiDAR point cloud. The method comprises: firstly, combining a region growing method based on normal and a region growing method based on distance, segmenting the big top surface from point cloud; then performing clustering to the remaining points, and segmenting the small top surface from each cluster by using a random sample consensus method; and by constantly improving the determination condition of angles between the top surfaces, and gradually segmenting the finer angle between the top surface. and finally, achieving automatic and accurate segmentation for various top surfaces of buildings, thereby laying a foundation for automatic modeling of three-dimensional buildings.

The invention relates to segmentation equipment, in particular to portable medicine segmentation equipment for the pharmacy department. The portable medicine segmentation equipment for the pharmacy department aims at achieving the technical purposes that tablets can be automatically segmented and segmentation is uniform. The portable medicine segmentation equipment for the pharmacy department comprises: an outer frame, wherein the number of the outer frame is one; a storage pipe, which is arranged in the middle of the upper part of the outer frame; observation plates, which are respectively arranged on the outer frame and the storage pipe; a rotating cover, which is rotationally arranged at the bottom of the outer frame; and a discharging mechanism, which is arranged on one inner side of the outer frame. According to the portable medicine segmentation equipment, the effects that tablets can be automatically cut, and segmentation is uniform are achieved; and tablets are discharged to a position above a supporting plate, and in the left-right moving process of a cutting knife, the tablets can be cut in a half-and-half mode, so the effect of evenly cutting the tablets can be achieved.

The invention relates to a mammary gland X-ray auxiliary diagnosis technology, and aims to provide a method for automatically segmenting mammary gland calcification points based on deep learning. Themethod comprises the steps of making a data set; pre-fetching the data; constructing a deep convolutional neural network; training a deep convolutional neural network by using the normalized data set;and reasoning a to-be-detected image by using the trained deep convolutional neural network. According to the method, by introducing related technologies of deep convolutional neural network learningand training, all calcification points on the mammary gland X-ray image can be rapidly and automatically segmented. Based on the application of the invention, the accuracy of judging canceration by adoctor can be improved.

The invention relates to the technical field of online quality detection of automatic production process and discloses a method for grading adhesive bulk agricultural products on line. The method comprises the following steps of: 1, establishing a normalized moment inertia (NMI) characteristic model and a weighting rotary inertia standard deviation model for qualified bulk agricultural products; and 2, grading bulk agricultural products to be graded on line by utilizing the NMI characteristic model and the weighting rotary inertia standard deviation model of the qualified bulk agricultural products and NMI characteristics and weighting rotary inertia standard deviation information of the bulk agricultural products to be graded. The NMI characteristics, the weighting rotary inertia standard deviation and a mathematical model are introduced into online grading of the bulk agricultural products, the adhesive bulk agricultural products are automatically and nondestructively cut according to the adhesion conditions of the bulk agricultural products on a production line by using the distance transformation and watershed method, and the adhesive bulk agricultural products are graded, so that the qualification rate is detected on line.

The embodiment of this specification discloses a segmentation method, device and equipment based on image data. The method includes: acquiring CTA image data to be processed; inputting the CTA image data to be processed into a segmentation model, segmenting the CTA image data to be processed, and obtaining three segmentation results, wherein the segmentation model It is a three-segmentation model obtained in advance through neural network training based on a semi-automatic labeling method. The three-segmentation results are the segmentation results of bone tissue, blood vessels and calcified tissue of the CTA image data to be processed. By adopting the segmentation method provided in this manual, the segmentation of bone tissue, calcified tissue and blood vessels can be automatically realized, which greatly saves analysis time and has high accuracy.

The embodiment of the invention discloses a segmentation method, device and equipment based on image data. The method comprises the following steps that CTA image data to be processed is acquired; the to-be-processed CTA image data are input into a segmentation model, the to-be-processed CTA image data are segmented, a three-segmentation result is obtained, the segmentation model is a three-segmentation model obtained in advance through neural network training on the basis of a semi-automatic labeling method, the three segmentation results are a bone tissue segmentation result, a blood vessel segmentation result and a calcified tissue segmentation result of the to-be-processed CTA image data. By adopting the segmentation method provided by the specification, the segmentation of bone tissues, calcified tissues and blood vessels can be automatically realized, the analysis time is greatly saved, and the accuracy is relatively high.

The invention discloses a method for segmenting ultrasound images of thyroid nodules with a cascaded fully convolutional neural network, comprising: constructing a simple fully convolutional neural network based on U-Net, and analyzing the thyroid nodules in ultrasound data according to the simple fully convolutional neural network Ultrasound images are segmented to segment the region of interest; the VGG19-FCN network is used as the downsampling layer to extract the deep features of the region of interest, so as to realize the automatic semantic segmentation of thyroid nodules; wherein, the simple full convolution neural network The network includes: five convolutional layers for downsampling, and five upsampling layers for upsampling; among them, the first five convolution convs are composed of two 3x3 convolutional layers and a pooling layer, Each convolutional layer uses ReLU as the activation function, and the last five upsampling layers are deconvolutional layers. The invention provides high-precision nodule images for the identification of benign and malignant thyroid nodules, thereby playing a better auxiliary role in medical diagnosis.

The present invention proposes an improved full convolutional neural network sequence image segmentation method for ceramic material parts, including steps: S10: manually labeling the collected original images, distinguishing objects and backgrounds with different categories, and obtaining training labels, Use the index mode to represent the label map of the training samples; S20: construct an improved network model based on the full convolutional neural network, and perform training; S30: calculate the loss function and backpropagation calculation loss function according to the gradient descent algorithm, and train the network Learning, the learning rate is reduced to a factor of 10 when the validation accuracy stops increasing. The fully convolutional neural network is an improved structure based on the convolutional neural network. On the basis of maintaining the good classification performance of CNN, it better maintains the spatial position relationship between pixel matrices, which is more conducive to global feature extraction, and can comprehensively Learning the visual characteristics of the object, it has good anti-interference ability, and can automatically separate the target object from the background to realize intelligent segmentation.

The invention discloses a sectional water feeding type solar water heater. A water storage tank (1) is connected with two water outlets, one water outlet is positioned in the top end of the water storage tank and connected with a water return pipe (4) through a pipeline, and a water outlet valve SV1 is arranged on the pipeline of the water outlet; the other water outlet is positioned in the middleof the water storage tank and connected with the water return pipe through a pipeline, and a water outlet valve SV2 is arranged on the pipeline of the water outlet; water of the water return pipe flows to a water bucket (2), and the top end of the water bucket is connected with a movable arm (5) used for controlling a travel switch SQ to be switched on and switched off; and the solar water heaterfurther comprises a thermistor Rt1, a photoresistor Rs1 and a peripheral circuit, wherein the thermistor Rt1 is mounted in the water storage tank and used for sensing the water temperature, the photoresistor Rs1 is used for sensing the illumination intensity, and the peripheral circuit is connected with the thermistor Rt1, the photoresistor Rs1, the water outlet valve SV1, the water outlet valveSV2, a water inlet valve SV3 and the travel switch SQ. According to the solar water heater, the water feeding level can be automatically selected according to the illumination intensity of sunlight and the water temperature condition.

The invention discloses a splitting method of a CD image of a solid engine. The method comprises the following steps: a CD faultage image data of the solid engine is read in, an artifact which is formed outside the solid engine owning to the air agitation formed in the CD detection is eliminated; the image, the artifact of which is eliminated, filters and eliminates background noise by median filter, meanwhile, the edges of each component of the solid engine are sharpened in the CD image and a casing, a propelling agent and an astropyle of the solid engine is cut in the CD image; a defect is cut from the propelling agent. By adopting the method, an outer layer air ring artifact, the casing, the propelling agent, the astropyle and various defects of the solid engine in the CD image are cutautomatically and accurately, thus solving the problem of low efficiency and poor accuracy in the CD detection image of the solid engine by the manual interpretation as well as being not limited by the eigenvalue of image information.

The invention provides a division method and device of left ventricular myocardium. The method comprises the following steps of: (1) positioning a left ventricular central point, and acquiring a region of interest in an image at a layer where the left ventricular central point is located; (2) positioning a blood pool region in the region of interest in the image at the layer where the left ventricular central point is located according to grey scale information and the left ventricular central point; (3) determining a mass center of the blood pool region, determining blood pool regions in an original image in all layers except the layer where the left ventricular central point is located, thereby obtaining a division result of left ventricular endocardium; and (4) obtaining a division result of left ventricular adventitia by a dynamic planning method according to the division result of the left ventricular endocardium, and thus obtaining a division result of the left ventricular myocardium. By the technical scheme provided by the invention, the left ventricular myocardium can be fully, uniformly and stably divided.

The invention discloses an image processing method applied to surgical navigation. The method comprises the steps that a tested image is acquired; if the acquired tested image comprises a multimodality image of a CT image and / or an MR image, multimodality image registration is performed; voxels of the tested image are segmented and marked; features of background, vessels, nerves, skeletons, foci and visceral organs where the foci are located are extracted, and feature vector preprocessing is performed; feature parameters of focus areas obtained after image segmentation are calculated; the end point of a surgical navigation path is determined though feature parameter calculation, a starting point area of the surgical navigation path is determined in a man-machine interaction mode, and path optimization is done. Through multimodality image registration and analysis, automatic and accurate segmentation of the foci, calculation of the parameters of the foci and output of an optimized path for cooperation with visual operation during surgical navigation are realized.

An image processing method for converting 2D (2-dimensional) video into 3D (3-dimensional) video comprises the steps of taking images, shot from different angles, of a target object as sample images, calculating color invariable images of the sample images and a to-be-identified image, extracting feature points from the color invariable images, matching the feature points in the color invariable image of the to-be-identified image with the feature points in the color invariable images of the sample images, judging whether the target object exists in the to-be-identified image according to distribution of the feature points matched successfully, if the target object exists in the to-be-identified image, dividing a target object outline based on the feature points matched successfully. The method can identify the target object when the 2D video is converted into the 3D video, precisely divides the object outline, and greatly reduces the workload of manual operation.