57 results about "Skeletonization" patented technology

The invention provides method, system and device for determining trabecular bone structure and strength by digital topological analysis, and offers, for the first time, a demonstration of superior associations between vertebral deformity and a number of architectural indices measured in the distal radius, thus permitting reliable and noninvasive detection and determination of the pathogenesis of osteoporosis. A preferred embodiment provides imaging in three dimension of a region of trabecular bone, after which the 3D image is converted into a skeletonized surface representation. Digital topological analysis is applied to the converted image, and each image voxel is identified and classified as a curve, a surface, or a junction; and then associated with microarchitectural indices of trabecular bone to quantitatively characterize the trabecular bone network. The invention is applicable in vivo, particularly on human subjects, or ex vivo.

Method for analysis of hydrocarbon potential of subterranean regions by generating surfaces or geobodies and analyzing them for hydrocarbon indications. Reflection-based surfaces may be automatically created in a topologically consistent manner where individual surfaces do not overlap themselves and sets of multiple surfaces are consistent with stratigraphic superposition principles. Initial surfaces are picked from the seismic data (41), then broken into smaller parts (“patches”) that are predominantly topologically consistent (42), whereupon neighboring patches are merged in a topologically consistent way (43) to form a set of surfaces that are extensive and consistent (“skeleton”). Surfaces or geobodies thus extracted may be automatically analyzed and rated (214) based on a selected measure (213) such as one or more direct hydrocarbon indications (“DHI”), e.g. AVO classification. Topological consistency for one or more surfaces may be defined as no self overlap plus local and global consistency among multiple surfaces (52).

The invention discloses a high-speed weak target flight path detection method of an image field, relating to the field of moving object tracking, and mainly solving the problem of extracting a moving object flight path and a target kinematic parameter from data with a low signal-to-noise ratio. The detection process is as follows: firstly, obtaining a time-Doppler image by mixing and accumulating target multiframe data; carrying out edge extracting on the basis of phase equalization model; carrying out skeletonization and binary image de-noising treatment on the edge detection result with a morphology method to realize edge center positioning as well as noise inhibition; and finally, using a least square method to detect the target flight path and estimate a target parameter. The invention can accurately estimate the advantages of a curve flight path parameter, and can be used for the data processing field for monitoring systems, such as radars and the like to realize high-speed weak target detection and tracking.

The invention discloses a skeletonization and template matching-based traffic police gesture identification method, which comprises the following steps of: (1) extracting key gestures of a traffic police, performing binarization on gesture sequence images, and normalizing the size of the binary images; (2) extracting the image outline of the key gestures of the traffic police, performing skeletonization by adopting outline deletion to obtain skeletons of the key gesture images of the traffic police, and storing the skeletons corresponding to various key gestures of the traffic police in a template library; and (3) extracting the key image of the current traffic police gesture sequence, performing skeletonization by adopting outline deletion to obtain the current gesture skeleton, matching the current gesture skeleton and each skeleton in the template library to find the most similar skeleton serving as the current traffic police gesture skeleton, and obtaining identification results of traffic police fingers. By the method, high reliability of the identification results can be ensured under low calculation complexity.

Technologies are provided for generating three-dimensional (3D) skeletons of target objects using images of the target objects captured from different viewpoints. Images of an object (such as a person) can be captured from different camera angles. Feature keypoints of the object can be identified in the captured images. Keypoints that identify a same feature in separate images can be correlated using truncated epipolar lines. For example, depth information for a keypoint can be used to truncate an epipolar line that is created using the keypoint. The correlated feature keypoints can be used to create 3D feature coordinates for the associated features of the object. A 3D skeleton can be generated using the 3D feature coordinates. One or more 3D models can be mapped to the 3D skeleton and rendered. The rendered one or more 3D models can be displayed on one or more display devices.

A computer-implemented method for determining a joint space width includes providing image data for a skeleton, thresholding the image data, and performing a connected component analysis on thresholded image data. The method further includes extracting contours of the thresholded image data according to the connected component analysis, performing a skeletonization of the thresholded image data using a first fast marching analysis of the thresholded image data, locating at least one finger joint of skeletonized image data, extracting bone boundaries using a second fast marching analysis of gradient information of the image data inside a region of interest, which includes a finger joint of the at least one finger joint, determining the joint space width given extracted bone boundaries, and outputting the joint space width.

In a method and apparatus for fully automatic detection of anomalies in vessel structures, a 3D volume data set of an imaging 3D measurement of the vessel structure is obtained and, for an evaluation device, the vessel structure is detected in the 3D volume data set and subsequently is skeletonized in order to obtain a three-dimensional course of skeletonization paths. Characteristic quantities of the vessel structure are automatically determined along the skeletonization paths as features that are significant for an anomaly to be detected, in order obtain one or more feature series. The determined feature series are classified by non-linear imaging and comparison with reference feature series that have been determined for different classes of known vessel structures that contain different anomalies and known vessel structures without anomalies. Anomalies corresponding to the classification thus are identified.

A method, system, apparatus, and computer program product is presented for management of a distributed data processing system. A network management framework dynamically adapts database operations so as to maximize system performance. A skeletonization mechanism is presented in which some objects are merely represented in a database cache as a skeleton object. A skeleton object holds only a fraction of an object's full complement of data; the skeleton object remains associated with its complete object, also termed its corresponding full object or non-skeleton object, which remains stored within its original location within a database. Skeletonized data can be updated on an as-needed or on-demand basis or when otherwise convenient; at that time, a skeleton handler can detect when a request is made for the object, retrieve the remaining data from persistent storage, and then mark the cached object as being a full object or non-skeleton object.

A method and system for reduction of a network topology-based system having automated optimization features is provided. The method allows for skeletonization to be performed by one or more desired processes including Data Scrubbing, Parallel and Series Pipe Removal techniques, and Branch Trimming techniques. The Data Scrubbing technique includes a loop retaining sensitivity setting that allows the user to determine how large the hydraulic loops are that remain in the system. The data scrubbing technique also includes a network-walking feature in which the software will check each element, when desired, and determine whether it is safe to remove that element without affecting the topological and hydraulic connectivity of the engineering model being reduced. The feature also includes a genetic algorithm-based feature that allows for hydraulics or system behavior to be reintroduced into the model, which may have been eliminated when the elements that gave rise to those effects were removed using the data scrubbing skeletonization technique.

The invention discloses a method for detecting self-adaptive growth of a plurality of weak target tracks in an image domain, which relates to the field of tracking of moving targets and mainly solves the problem that the existing method is difficult to detect a plurality of moving target curve tracks under the condition of low signal-to-noise ratio. The method comprises the following detection processes: firstly, on the basis of processing radar echoes to acquire a distance-time image, extracting edges based on a phase consistency model; then, by adopting a digital image processing method for the edge detection result, realizing the purposes of positioning edge centers, removing fake edges and communicating the connected regions with the consistent edge trend through skeletonization and self-adaptive region growth technology; and finally, estimating target moving tracks and moving parameters by combining a clearing method and a iterative least squares method. The invention has the advantage of being capable of accurately estimating the track parameters of a plurality of curves, and can be applied to the data processing field of monitoring systems such as radars and the like to realize high-speed weak target detection and tracking.

Disclosed is a method for measuring trabecular bone parameters from MRI images, including: scanning an experimental group with a 3D MRI scanner; segmenting the MRI images to extract bone area and perform skeletonization of the bone area; detecting end-point, joint and branch voxels in the skeleton to analyze bone structure; and measuring trabecular bone parameters based on the result of the structural analysis. The method enables diagnosing osteoporosis.

The embodiment of the invention provides a crop seedling condition and seedling potential acquisition and analysis method and device, and the method comprises the steps of obtaining an aerial image ofa target area, carrying out the ridge cell recognition and background segmentation of the aerial image, and obtaining a plant image without a background; rotating the plant image to a ridge horizontal direction, and searching a plant contour based on a preset area threshold to obtain a plant contour of each plant; traversing the outline of each plant, and carrying out skeletonization treatment respectively to obtain a stem part of each plant; and sorting and clustering the stalk parts based on the ridge direction to obtain stalk positions, judging whether the two adjacent plants are lack of seedlings or not based on the distance between the stalk positions of every two adjacent plants, and obtaining the number of the lack of seedlings if the two adjacent plants are lack of seedlings. Themethod has the advantages that the seedling condition seedling potential phenotype parameters of crops can be acquired in real time according to aerial images, the seedling condition seedling potential can be analyzed and judged, the multi-task and multi-site simultaneous analysis can be performed, the seedling period characters of crop growth can be accurately and quickly positioned, the cost islow, and the flexibility is high.

The invention relates to a method for constructing a dermatoglyph classification prediction model by introducing a ResNet deep learning network. The method comprises steps of using an intelligent terminal device for fully collecting a sample dermatoglyph original image and sequentially carrying out normalization, Wiener filtering denoising, Sobel operator algorithm sharpening, binarization algorithm processing and OPTA pixel skeletonization processing; repairing and enhancing by adopting a GAN generative adversarial network model algorithm, and manually labeling each sample dermatoglyph image;and finally, establishing a dermatoglyph classification prediction model, optimizing a loss function, iteratively training the model, and verifying to obtain a dermatoglyph classification model. According to the method, a CNN-based ResNet deep learning network is introduced to construct a dermatoglyph classification prediction model; when the constructed model is applied, different dermatoglyph feature images can be learned and analyzed from the aspects of multiple dimensions and multiple features, more features are extracted from dermatoglyph image information, and high accuracy is achievedin dermatoglyph recognition and classification.

The invention discloses a finger vein recognition unlocking method and system. According to the method, binaryzation is carried out on a finger vein image in an image preprocessing stage; a region ofinterest is denoised and extracted; operations such as enhancement and normalization are performed; a region of interest is accurately extracted; de-noising is carried out again after image segmentation; cavities are filled. image post-processing such as burr filtering and skeletonization is carried out; According to the method, the finger vein to-be-identified sample with clear and effective finger vein characteristics can be quickly and efficiently extracted, the finger vein to-be-identified sample is quickly matched with the finger vein identification template stored in the database, if thematching is successful, the unlocking signal is output for unlocking, and the finger vein identification and unlocking efficiency is improved.

A morphological inspection method based on a comparison of real images—a real reference image and a real inspected image of an inspected object is disclosed.

The embodiment of the invention discloses an image segmentation method and device, equipment and a storage medium. The method comprises the steps of obtaining image data of a target object, and determining a to-be-segmented image according to the image data; Segmenting the to-be-segmented image based on a preset low-threshold region growing algorithm to obtain an initial segmented image with a tubular structure; performing skeletonization processing on the initial segmented image to obtain an image tree with a tubular structure; and cutting the image tree based on a preset high-threshold probability cutting algorithm, and determining a segmented image of the tubular structure according to the cut image tree. According to the technical scheme provided by the embodiment of the invention, theproblem of incomplete segmentation or segmentation leakage in the existing scheme is solved, the tubular structures such as trachea, blood vessels and coronary vessels in the medical image can be accurately segmented, and a basis is provided for medical diagnosis.

The invention relates to an image skeletonization method of a printing circuit board based on FPGA which firstly inputs an image to analyze the neighborhood of each pixel and count the number n with a pixel value of 1 among the surrounding 8 pixels and the pixel with a switch time S between 0 and 1, with an output being 1 and a preservation being 1; then, judging whether a central pixel is 1; deleting the central pixel if the central pixel is zero; when n is more than 1 but is equal to or less than 6, S is equal to 2, and circulating time is odd, investigating the pixels on the right and lower directions of the central pixel and the pixels on the left and right directions of the central pixel, judging whether 0 and 1 are cross, outputting 0 and deleting the pixel; when n is more than 1 but is equal to or less than 6, S is equal to 4, n is equal to 4, and the circulating time is odd, investigating the pixels on the upper-left and upper-right directions of the central pixel and the pixels on the lower-left and lower-right directions of the central pixel; judging whether the lower side pixel of the central pixel is 1 and whether 0 and 1 are two communicating areas; if the pixel is 0, deleting the pixel. The invention can fast and accurately withdraw the image skeleton.

The invention relates to a determination method for quantitatively detecting a thick cabo rate and long and short cabo rates in caboes on the basis of an X-ray transmission image. By utilizing X-ray image feature difference between the caboes and tobacco leaves, the method adopts gray-scale morphological denoising and image segmentation by a region growing method to preprocess an image, meanwhile,a fuzzy C-mean value clustering algorithm based on an unsupervised machine learning function is adopted for belonging judgment, and according to a shape judgment factor, an image recognition algorithm for the nondestructive detection and recognition of caboes in tobacco leaf material are designed and implemented. Moreover, according to a conventional definition standard for thick caboes and longand short thick caboes, a quantitative detection algorithm for cabo morphology is further established, which includes the detection and calculation of the thick cabo rate and the long and short cabo rates, a thick cabo rate detection algorithm mainly relates to cabo diameter extracted by a segmented bounding rectangle method, thickness cabo judgment and thick cabo quality calculation. A long and short cabo rate detection algorithm mainly relates to cabo skeletonization, skeleton length calculation, long and short cabo judgment and quality calculation. The advantage is that the determination method can remarkably increase the accuracy of a detection result and working efficiency and eliminate anthropogenic influence.

The invention relates to the technical field of medical image processing and particularly relates to a cell counting method based on skeleton extraction, which comprises the following steps: S1, obtaining a to-be-processed histopathology image, carrying out graying processing on each to-be-processed image, and then carrying out image segmentation on each grayed image to obtain a cell binary imageof each to-be-processed image; S2, performing morphological processing on each cell binary image to obtain a cell binary image with intracellular hole filling and cell edge noise and impurity noise point removal; And S3, performing cell skeletonization processing on each morphologically processed cell binary image to obtain the number of cell skeletonization nodes in each image, and obtaining a cell counting result of each to-be-processed image according to the relationship between the number of cell skeletonization nodes in each image and the number of cells. The obtained cell counting resultis good in accuracy, and the method is high in operation efficiency.

The invention discloses an image identification system and an image identification method. The system comprises an image extraction unit, a processing module, a storage unit and an identification module. The image extraction unit performs image extraction operation at a selected position; the processing module is connected with the image extraction unit, processes the images extracted by the image extraction unit by skeletonization, and acquires the outline characteristics of the skeletonized images; the storage unit is connected with the processing module, and stores the outline characteristics of the skeletonized images; and the identification module is connected with the storage unit, and compares the outline characteristics of the skeletonized images acquired at different time to identify whether the images acquired at different time are matched.

The invention belongs to the field of computer vision and pattern recognition, particularly relates to a Chinese character image stroke extraction method and system based on a full convolutional neural network, and aims to solve the problem that handwriting character strokes written freely are difficult to extract. The Chinese character image stroke extraction method comprises the following steps:carrying out region extraction on an acquired Chinese character image; carrying out skeletonization operation on the overlapped area and the non-overlapped area; calculating the coherence between anystroke segments of the overlapped area after skeletonization; and connecting the stroke segments belonging to the same stroke in the overlapped area, and combining the stroke segments directly connected in the non-overlapped area into a complete skeleton form stroke. On one hand, stroke extraction of handwritten Chinese characters can still be achieved under the condition that strokes of the handwritten Chinese characters which are written freely are overlapped; and on the other hand, a character synthesis method is adopted for obtaining a training sample, and different marking information ofthe training sample in different tasks is attached, so that the labor cost is greatly saved.

A system for plant leaf identification includes: a plant image capturing unit which captures an image of a target plant to generate a plant image; a plant area image extraction unit which separates a background area and a plant area in the plant image to generate a plant area image including the plant area; a plant area image skeletonization unit which skeletonizes the plant area image to generate a skeletonized plant area image; a candidate leaf path generation unit which identifies a root vertex, a junction vertex and a leaf tip vertex in the skeletonized plant area image, and generates a plurality of candidate leaf paths by calculating all possible paths from the root vertex to the leaf tip vertex; and a final leaf path reconstruction unit which reconstructs a final leaf path matching the plant image by selecting the plurality of candidate leaf paths.

A method and system for optimization of a simplified engineering model is provided. A software tool efficiently simplifies an engineering model such as a water distribution system and preserves the hydraulic accuracy of the simplified model. The system includes a software program that employs a genetic algorithm to evolve solutions for reinstating the behavior of the original network into a simplified network. The genetic algorithm can be used for identifying the less sensitive hydraulic elements (links and nodes), and removing them or replacing them with the best-fit element parameters produced by the genetic algorithm module of the present invention. A element-by-element skeletonization approach generates the layout of a skeleton network and then identifies potential equivalent elements for replacing series pipes and loops. Yet another alternative includes taking a skeleton network model, and using the techniques of the present invention, optimizes the demand distribution and pipe conductance of the skeletonized model.