61 results about "Recursive segmentation" patented technology

The invention provides an AI-based heart failure risk assessment method and system, and relates to the technical field of artificial intelligence, and the method comprises the steps: recognizing a mutation point of a physiological index time sequence through a recursive segmentation algorithm, extracting a mutation feature matrix, carrying out the time sequence mapping, recognizing a periodic mutation and gradual change mode, and converting into a risk score to generate a risk accumulation curve. Calculating index conduction time delay, drawing a risk conduction path diagram, marking an intervention time window, and finally determining a heart failure risk level and generating an evaluation report. According to the invention, early accurate recognition of the heart failure risk can be realized, and the best opportunity is provided for clinical intervention.

Disclosed are methods and apparatuses for decoding an image. A method includes receiving a bitstream obtained by encoding the image; dividing a first coding block into a plurality of second coding blocks; generating a prediction block of a second coding block based on syntax information obtained from the bitstream; and reconstructing the second coding block based on the prediction block and a residual block of the second coding block, the residual block being obtained by performing a dequantization and an inverse-transform on quantized transform coefficients from the bitstream. The first coding block has a recursive division structure. The first coding block is divided based on at least one of a quad tree division, a binary tree division or a triple tree division.

Disclosed are methods and apparatuses for decoding an image. A method includes receiving a bitstream obtained by encoding the image; dividing a first coding block into a plurality of second coding blocks; generating a prediction block of a second coding block based on syntax information obtained from the bitstream; and reconstructing the second coding block based on the prediction block and a residual block of the second coding block, the residual block being obtained by performing a dequantization and an inverse-transform on quantized transform coefficients from the bitstream. The first coding block has a recursive division structure. The first coding block is divided based on at least one of a quad tree division, a binary tree division or a triple tree division.

A system and method for 3D anomaly detection and tracking are provided that uses multimodal fusion, reduced training data, recursive segmentation and histogram statistic distance. The anomaly may be a defect or a configuration error.

The invention discloses a distribution line equipment parameter calculation method using a three-dimensional point cloud, and the method comprises the following steps: employing advanced visible light shooting equipment or other three-dimensional sensing equipment to obtain the point cloud data of distribution line equipment and the surrounding environment of the distribution line equipment; the detailed geometrical shape and spatial distribution of the distribution line equipment are obtained through the point cloud data, and a high-quality data source is provided for subsequent equipment parameter extraction; after the point cloud data are obtained, the point cloud data are preprocessed so as to improve the quality of the point cloud data. Three-dimensional point cloud data are collected through visible light shooting equipment and a high-precision camera, and high-precision extraction of equipment parameters is realized in combination with principal component analysis and an optimization algorithm. The measurement efficiency is improved through automatic processing of point cloud data, parameters are quickly calculated by adopting an octree and recursive segmentation algorithm for complex equipment, and the method has relatively high adaptability and anti-noise capability and is suitable for various detection scenes.

Disclosed are methods and apparatuses for decoding an image. A method includes receiving a bitstream obtained by encoding the image; dividing a first coding block into a plurality of second coding blocks; generating a prediction block of a second coding block based on syntax information obtained from the bitstream; and reconstructing the second coding block based on the prediction block and a residual block of the second coding block, the residual block being obtained by performing a dequantization and an inverse-transform on quantized transform coefficients from the bitstream. The first coding block has a recursive division structure. The first coding block is divided based on at least one of a quad tree division, a binary tree division or a triple tree division.

The invention relates to the technical field of three-dimensional point cloud segmentation and terrain feature extraction, and discloses a step terrain segmentation method for enhancing L0 gradient minimization, which comprises the following steps of: calculating point cloud features based on iterative principal component analysis: constructing a dynamic weight distribution mechanism by introducing a robust M estimator; direction self-adaptive neighborhood optimization is carried out on the slope and panel transition area of the original point cloud, the continuity of the normal vector of the slope area is enhanced, and the geometric difference of the transition area is highlighted; multi-scale super voxel segmentation: adopting a normalized spatial metric segmentation method, generating super voxel units by using a distance, normal vector and perpendicularity weighted normalization mechanism, performing planar and non-planar classification on super voxels in combination with a multi-scale strategy, and performing recursive segmentation on non-planar super voxels at a smaller resolution; performing curved surface segmentation based on L0 gradient minimization and global energy optimization; according to the method, the definition of the segmentation boundary and the completeness of slope structure extraction can be remarkably improved, and the slope and the disc table can be accurately distinguished by using the vertical characteristics.

The invention discloses an image compression method based on adaptive quadtree segmentation and standard deviation optimization, and provides an efficient compression scheme driven by dynamic statistics aiming at the problems of over-segmentation, low calculation efficiency and detail loss under a high compression rate in a traditional quadtree compression algorithm. Specifically, the compression ratio and the reconstruction quality are effectively balanced by calculating the RGB channel standard deviation of an image region in real time and combining region area weighted income analysis to adaptively control the recursive segmentation depth; an area weighted error constraint mechanism is introduced, segmentation of a large-size low-variance region is terminated preferentially, and the redundant data volume is remarkably reduced; a standard deviation-driven priority ranking strategy is adopted, fine processing is performed on high-complexity sub-regions, and rapid approximation is realized by combining mean color filling, so that the algorithm efficiency is optimized while the peak signal-to-noise ratio is ensured. Through the synergistic effect of the technologies, on the premise that high-frequency details of the image are reserved, calculation complexity and storage requirements are greatly reduced, the method is suitable for real-time image transmission, mobile terminal processing and low-bandwidth communication scenes, and a reliable solution is provided for high-precision and low-delay compression requirements.

Disclosed are methods and apparatuses for decoding an image. A method includes receiving a bitstream obtained by encoding the image; dividing a first coding block into a plurality of second coding blocks; generating a prediction block of a second coding block based on syntax information obtained from the bitstream; and reconstructing the second coding block based on the prediction block and a residual block of the second coding block, the residual block being obtained by performing a dequantization and an inverse-transform on quantized transform coefficients from the bitstream. The first coding block has a recursive division structure. The first coding block is divided based on at least one of a quad tree division, a binary tree division or a triple tree division.

Disclosed are methods and apparatuses for decoding an image. A method includes receiving a bitstream obtained by encoding the image; dividing a first coding block into a plurality of second coding blocks; generating a prediction block of a second coding block based on syntax information obtained from the bitstream; and reconstructing the second coding block based on the prediction block and a residual block of the second coding block, the residual block being obtained by performing a dequantization and an inverse-transform on quantized transform coefficients from the bitstream. The first coding block has a recursive division structure. The first coding block is divided based on at least one of a quad tree division, a binary tree division or a triple tree division.

Disclosed are methods and apparatuses for decoding an image. A method includes receiving a bitstream obtained by encoding the image; dividing a first coding block into a plurality of second coding blocks; generating a prediction block of a second coding block based on syntax information obtained from the bitstream; and reconstructing the second coding block based on the prediction block and a residual block of the second coding block, the residual block being obtained by performing a dequantization and an inverse-transform on quantized transform coefficients from the bitstream. The first coding block has a recursive division structure. The first coding block is divided based on at least one of a quad tree division, a binary tree division or a triple tree division.

Disclosed are methods and apparatuses for decoding an image. A method includes receiving a bitstream obtained by encoding the image; dividing a first coding block into a plurality of second coding blocks; generating a prediction block of a second coding block based on syntax information obtained from the bitstream; and reconstructing the second coding block based on the prediction block and a residual block of the second coding block, the residual block being obtained by performing a dequantization and an inverse-transform on quantized transform coefficients from the bitstream. The first coding block has a recursive division structure. The first coding block is divided based on at least one of a quad tree division, a binary tree division or a triple tree division.

The present application relates to a method and device for detecting the position of lamp beads, as well as equipment and products, the method comprising: determining the serial number of each lamp bead according to the total number of lamp beads in a lamp chain; recursively segmenting the lamp chain according to the serial number to obtain multiple segments, and controlling each segment obtained by the current segmentation to present different apparent colors with other segments through lamp beads after each segmentation; identifying the spatial position and apparent color of the luminous lamp beads in the real-shot image of the lamp chain collected corresponding to each segmentation, and determining the serial number range to which the luminous lamp beads in the image belong according to the segment to which the apparent color belongs; determining the mapping relationship between the spatial position and serial number of each lamp bead according to the spatial position relationship between adjacent lamp beads in the real-shot image and the serial number relationship in the serial number range to which they belong. The present application improves the accuracy and efficiency of lamp bead position and serial number matching through a recursive segmentation method, and enhances the adaptability and scalability of the system in complex scenarios.

A block partitioner includes a quad splitter structured to partition a target block obtained by recursive partitioning in half in both a horizontal direction and a vertical direction to generate four blocks, and a binary / ternary splitter structured to partition the target block obtained by recursive partitioning into two or three in the horizontal direction or the vertical direction to generate two or three blocks, and the binary / ternary splitter disallows partitioning of the target block in the horizontal direction when partitioning of the target block in the horizontal direction causes the target block obtained by partitioning to be located beyond a right side of a picture boundary, and disallows partitioning of the target block in the vertical direction when partitioning of the target block in the vertical direction causes the target block obtained by partitioning to be located beyond a lower side of the picture boundary.

An image encoding device adapted to segment an image into blocks and encode the image in units of blocks resulting from segmenting the image is provided. A block segmentation unit recursively segments the image into rectangles of a predetermined size to generate a block subject to encoding. A bitstream generation unit encodes block segmentation information of the block subject to encoding. The block segmentation unit includes: a quartering unit that quarters a target block in recursive segmentation in a horizontal direction and a vertical direction to generate four blocks; and a halving unit that halves a target block in recursive segmentation in a horizontal or vertical direction to generate two blocks. When previous recursive segmentation is halving, the halving unit prohibits a target block subject to current recursive segmentation from being segmented in the same direction as a direction in which the block was segmented in the previous recursive segmentation.

The application belongs to the field of image communication, and provides a bidirectional dynamic encryption and decryption method and system for image and text based on geometric segmentation, which comprises the following steps: the server initially segments a target image, recursively segments each segmented image block, obtains segmented image sub-blocks, disarranges the segmented image sub-blocks in each segmented image block in disorder to obtain an encrypted target image; extracts an encrypted text key based on the encrypted target image; combines target user information, target text, an initial segmentation number of the target image and a disorder sequence of the encrypted target image to obtain combined information, encrypts the combined information by using the encrypted text key to obtain an encrypted target text; the front end receives the encrypted target image and the encrypted target text, obtains a decryption key based on the encrypted target image, decrypts the encrypted target text by using the decryption key to obtain the combined information; decrypts the encrypted target image based on the combined information to obtain a target image; and the network transmission of the image and the text is more secure.

Disclosed are methods and apparatuses for decoding an image. A method includes receiving a bitstream obtained by encoding the image; dividing a first coding block into a plurality of second coding blocks; generating a prediction block of a second coding block based on syntax information obtained from the bitstream; and reconstructing the second coding block based on the prediction block and a residual block of the second coding block, the residual block being obtained by performing a dequantization and an inverse-transform on quantized transform coefficients from the bitstream. The first coding block has a recursive division structure. The first coding block is divided based on at least one of a quad tree division, a binary tree division or a triple tree division.

Disclosed are methods and apparatuses for decoding an image. A method includes receiving a bitstream obtained by encoding the image; dividing a first coding block into a plurality of second coding blocks; generating a prediction block of a second coding block based on syntax information obtained from the bitstream; and reconstructing the second coding block based on the prediction block and a residual block of the second coding block, the residual block being obtained by performing a dequantization and an inverse-transform on quantized transform coefficients from the bitstream. The first coding block has a recursive division structure. The first coding block is divided based on at least one of a quad tree division, a binary tree division or a triple tree division.

Disclosed are methods and apparatuses for decoding an image. A method includes receiving a bitstream obtained by encoding the image; dividing a first coding block into a plurality of second coding blocks; generating a prediction block of a second coding block based on syntax information obtained from the bitstream; and reconstructing the second coding block based on the prediction block and a residual block of the second coding block, the residual block being obtained by performing a dequantization and an inverse-transform on quantized transform coefficients from the bitstream. The first coding block has a recursive division structure. The first coding block is divided based on at least one of a quad tree division, a binary tree division or a triple tree division.

To provide an image encoding apparatus, method, and program that improve encoding efficiency by performing block division suitable for image encoding and decoding.SOLUTION: There is provided an image encoding apparatus 100 that divides an image into blocks and performs encoding in a divided-block unit. A block division unit 101 recursively divides an image into rectangular shapes of a predetermined size to generate blocks to be encoded. An encoded bit string generation unit 105 encodes block division information of each block to be encoded. The block division unit further includes: a four-way division part that divides a target block in recursive division into four pieces horizontally and vertically to generate four blocks; and a two-way division part that divides a target block in recursive division into two pieces horizontally or vertically to generate two blocks. When the previous recursive division was two-way division, the two-way division part prohibits dividing the target block in the current recursive division in the same direction as the direction in which the block was divided in the previous recursive division.SELECTED DRAWING: Figure 1

Disclosed are methods and apparatuses for decoding an image. A method includes receiving a bitstream obtained by encoding the image; dividing a first coding block into a plurality of second coding blocks; generating a prediction block of a second coding block based on syntax information obtained from the bitstream; and reconstructing the second coding block based on the prediction block and a residual block of the second coding block, the residual block being obtained by performing a dequantization and an inverse-transform on quantized transform coefficients from the bitstream. The first coding block has a recursive division structure. The first coding block is divided based on at least one of a quad tree division, a binary tree division or a triple tree division.

The application discloses an inverted index iterative self-index optimization method based on a Top-K query model, first, an initial inverted index is constructed with the aid of an open source tool, which is converted into an equal word compression format and reorganized into a block interleaved storage structure; subsequently, based on the Top-K query model, a multi-layer iterative self-index is constructed through recursive segmentation and sampling; further, a query performance model containing parameters such as disk I / O and decompression speed is established, the relationship between query benefits and storage overhead is quantified, and the optimal synchronous distance and the number of iterative layers are solved; finally, fast random access of a mass disk compressed inverted index is realized. Through the variable-length synchronous distance design, the Top-K model parameter optimization and the multi-layer iterative index structure, the application significantly improves the query efficiency, reduces the data reading and decompression overhead, can maintain the Pareto optimal balance of the query performance and the storage efficiency in different hardware environments, and is suitable for a mass data high-concurrency query scene.

This invention provides an encoding method and apparatus for images by recursively dividing them into leaf blocks using a multi-tree partitioning method. [Solution] The method includes the steps of: determining whether the sum of the x-coordinate position and width of a given block is greater than the picture width, and determining whether the sum of the y-coordinate position and height of a given block is greater than the picture height, and determining a set of available split modes for a given block that extends beyond the boundary of an image; and selecting from the set of available split modes at least one of four divisions or two divisions in a vertical direction or two divisions in a horizontal direction, which divides the given block into two subblocks along a split line that is parallel to the boundary of the image within the given block and at least one of which has a split line that is juxtaposed with the boundary of the image and is centered inside the given block.

The invention discloses an intelligent Excel data segmentation method and system based on an AI desktop cloud customer service assistant, and belongs to the technical field of artificial intelligence. Line-by-line labeling of an Excel header is achieved through a structured semantic labeling engine, and intelligent splitting of the situation of merging cells is supported; intelligent segmentation which is compatible with an expected segmentation size and accords with semantics is carried out on the preprocessed text through a self-adaptive segmentation engine; the structured semantic annotation engine is used for loading an Excel document, splitting and merging cells, carrying out null value filling, and generating a semantic annotation text for each row of data in combination with a header; according to the self-adaptive segmentation engine, a recursive segmentation algorithm is adopted for the marked text, all paragraphs are kept together as long as possible on the premise that the preset maximum paragraph length is not exceeded, and the semantic context is richer. According to the method, the question and answer accuracy of the AI desktop cloud customer service assistant can be remarkably improved.

The invention belongs to the technical field of information, and particularly relates to a sliding fence product quantification method and device. The method comprises the following steps: obtaining N input D-dimensional vectors which are regarded as an N * D matrix, and carrying out initial segmentation to obtain a small-value matrix, a large-value matrix and an indication matrix; then recursive segmentation is carried out, a small-value matrix and a large-value matrix which are obtained after recursive segmentation are combined into a new matrix, and subspaces are divided in a sliding window mode; k-means clustering is executed in each subspace, and a cluster center is used as a codebook to encode all vectors in the subspace; and obtaining a final code of each element through a Count-min operation, and combining the final codes to obtain a quantization result. According to the method, improvement and optimization of a product quantization algorithm are realized, the memory overhead can be effectively reduced on the premise of ensuring the precision, and the method can be widely applied to the fields of large language model weight quantization, vector database management, KV cache optimization, image compression, image compression and the like which need efficient matrix quantization.

Disclosed are methods and apparatuses for decoding an image. A method includes receiving a bitstream obtained by encoding the image; dividing a first coding block into a plurality of second coding blocks; generating a prediction block of a second coding block based on syntax information obtained from the bitstream; and reconstructing the second coding block based on the prediction block and a residual block of the second coding block, the residual block being obtained by performing a dequantization and an inverse-transform on quantized transform coefficients from the bitstream. The first coding block has a recursive division structure. The first coding block is divided based on at least one of a quad tree division, a binary tree division or a triple tree division.

The invention relates to an aerial image slope vegetation coverage state evaluation method and system, and belongs to the technical field of image processing, and the method comprises the steps: collecting a slope image through an unmanned plane, calculating a green index, classifying pixel points, and recognizing a green plant connected domain; a suspected shadow area is extracted through recursive segmentation and brightness comparison, and the green plant structure complexity is calculated; evaluating the vegetation health degree based on the RGB channel relative value; dividing grids to analyze vegetation distribution uniformity; and finally, fusing four indexes of coverage rate, structural complexity, health degree and distribution uniformity to generate a slope vegetation coverage quantization parameter, and comparing with a threshold value to judge whether the vegetation coverage state is good or not, so that the slope vegetation coverage evaluation is more accurate.

Disclosed are methods and apparatuses for decoding an image. A method includes receiving a bitstream obtained by encoding the image; dividing a first coding block into a plurality of second coding blocks; generating a prediction block of a second coding block based on syntax information obtained from the bitstream; and reconstructing the second coding block based on the prediction block and a residual block of the second coding block, the residual block being obtained by performing a dequantization and an inverse-transform on quantized transform coefficients from the bitstream. The first coding block has a recursive division structure. The first coding block is divided based on at least one of a quad tree division, a binary tree division or a triple tree division.

This invention relates to a method and system for assessing the vegetation cover status of slopes using aerial imagery, belonging to the field of image processing technology. The method includes: acquiring slope images using a drone; calculating a green index and classifying pixels; identifying connected vegetation regions; extracting suspected shadow areas through recursive segmentation and brightness comparison; calculating the structural complexity of vegetation; assessing vegetation health based on relative values ​​of RGB channels; analyzing vegetation distribution uniformity by dividing the data into grids; and finally fusing four indicators—coverage, structural complexity, health, and distribution uniformity—to generate quantitative parameters for slope vegetation cover. These parameters are then compared with threshold values ​​to determine whether the vegetation cover status is good. This invention makes slope vegetation cover assessment more accurate.