Abstract: A method for level partition of a point cloud includes: decoding a point cloud bitstream, the point cloud bitstream including geometry information and attribute information of the point cloud; determining a maximum permissible value of a first syntax element in the point cloud bitstream; parsing the point cloud bitstream to determine a value of the first syntax element according to the maximum permissible value; determining, according to the value of the first syntax element, a number of levels of detail (LODs) in a process of decoding the attribute information; and decoding the attribute information according to the number of the LODs and the geometry information, to determine a reconstructed point cloud of the point cloud.

Abstract: A computer device obtains a to-be-segmented image that includes a plurality of extreme points corresponding to a target object in the to-be-segmented image. The device generates first image feature information according to the to-be-segmented image. The first image feature information includes N image matrices and a first heatmap that is generated according to the extreme points. The device obtains a first image segmentation region corresponding to the first image feature information using a first image segmentation model. The device also obtains a second heatmap according to an annotation point corresponding to the first image segmentation region and the first image segmentation region. The device obtains a second image segmentation region corresponding to the to-be-segmented image using a second image segmentation model. The device generates an image recognition result according to the second image segmentation region.

Abstract: Disclosed are an optical anti-counterfeiting element and an optical anti-counterfeiting product utilizing the same, the optical anti-counterfeiting element comprising: a substrate (12) including a first surface and a second surface opposite to each other; micro-sampling tools (11) on the first surface; and one or more micro graphic and text units formed on the second surface, wherein the micro graphic and text unit includes micro graphic and text strokes, the micro graphic and text strokes are arranged in a non-periodic and non-fixed-geometry manner, and when the second surface is observed through the micro-sampling tools, a stereoscopic animated anti-counterfeiting feature can be seen.

Abstract: An object is to provide an endoscopic image learning device, an endoscopic image learning method, an endoscopic image learning program, and an endoscopic image recognition device that appropriately learn a learning model for image recognition for recognizing an endoscopic image in which a treatment tool for an endoscope appears. The object is achieved by an endoscopic image learning device including an image generation unit and a machine learning unit. The image generation unit generates a superimposed image where a foreground image in which a treatment tool for an endoscope is extracted is superimposed on a background-endoscopic image serving as a background of the foreground image, and the machine learning unit performs the learning of a learning model for image recognition using the superimposed image.

Abstract: An image generation unit generates a contrast-reduced image in which contrast in a region other than a soft part is reduced in a radiographic image indicating a subject including the soft part and a bone part. A body thickness derivation unit derives the body thickness of the subject based on the contrast-reduced image.

Abstract: A system and method for generating augmented segmented image set obtain are provided. The method may include: obtaining a first image including a first anatomical structure of a first object; determining first feature data of the first anatomical structure; determining one or more first transformations related to the first anatomical structure, wherein a first transformation includes a transformation type and one or more transformation parameters related to the transformation type; applying the one or more first transformations to the first feature data of the first anatomical structure to generate second feature data of the first anatomical structure; and generating a second image based on the second feature data of the first anatomical structure.

Abstract: A multi-source domain adaptation model by aligning partial features includes a general feature extraction module, a feature selection module for partial feature extraction with the dedicated loss function, three partial feature alignment losses, and two classifiers for adversarial training, where the three partial feature alignment losses include an intra-class partial feature alignment loss, an inter-domain partial feature alignment loss, and an inter-class partial feature alignment loss. With the partial features extracted by the general feature extraction module and the feature selection module following three different partial feature alignment losses, the model is capable of clustering the samples from the identical categories and isolating samples from distinct classes.

Abstract: Disclosed is a system and associated methods for compressing motion within an animated point cloud. The resulting compressed file encodes different transforms that recreate the motion of different sets of points across different point clouds or frames of the animation in place of the data for the different sets of points from the different point clouds. The compression involves detecting a motion that changes positioning of a set of points between a first point cloud and subsequent point clouds of an uncompressed encoding of two or more frames of an animation. The compression further involves defining a transform that models the motion, and generating a compressed animated point cloud by encoding the data of the first point cloud in the compressed animated point cloud, and by replacing the data for the set of points in the one or more subsequent point clouds with the transform.

Abstract: The present disclosure relates to image processing apparatus and method that can prevent a reduction in image quality. Geometry data that is a frame image having arranged thereon a projected image obtained by projecting 3D data representing a three-dimensional structure on a two-dimensional plane and includes a special value indicating occupancy map information in a range is generated. The generated geometry data is encoded. Further, the encoded data on the geometry data is decoded, and a depth value indicating a position of the 3D data and the occupancy map information are extracted from the decoded geometry data. The present disclosure is applicable to, for example, an information processing apparatus, an image processing apparatus, electronic equipment, an information processing method, or a program.

Abstract: A system and method for fisheye image processing can be configured to: receive fisheye image data from at least one fisheye lens camera associated with an autonomous vehicle, the fisheye image data representing at least one fisheye image frame; partition the fisheye image frame into a plurality of image portions representing portions of the fisheye image frame; warp each of the plurality of image portions to map an arc of a camera projected view into a line corresponding to a mapped target view, the mapped target view being generally orthogonal to a line between a camera center and a center of the arc of the camera projected view; combine the plurality of warped image portions to form a combined resulting fisheye image data set representing recovered or distortion-reduced fisheye image data corresponding to the fisheye image frame; generate auto-calibration data representing a correspondence between pixels in the at least one fisheye image frame and corresponding pixels in the combined resulting fisheye image

Abstract: An image processing apparatus includes a correction unit, a first detection unit, a second detection unit, and a merging unit. The correction unit is configured to perform correction of a misalignment between a plurality of images captured at exposures different from each other. The first detection unit is configured to detect an inappropriate area from the plurality of images. The second detection unit is configured to detect a feature point for the correction by preferentially using an area other than the inappropriate area of the plurality of images. The merging unit is configured to merge the plurality of images on which the correction has been performed.

Abstract: A target cell marking method, including: determining an original image format of the original scanned image, and converting the original scanned image into a first image in a preset image format; segmenting the first image into a plurality of image blocks and recording arrangement positions of the image blocks in the first image; respectively inputting the image blocks into a preset deep learning detection model to obtain first position information of target cells in the image blocks; determining second position information of the target cells in the first image according to the first position information and the corresponding arrangement positions; integrating the image blocks according to the arrangement positions to obtain a second image, and marking the target cells in the second image; and converting the second image marked by the target cells into a third image in the original image format, and displaying the third image.

Abstract: The present invention provides an image processing circuit including a receiving circuit, a reference value calculating circuit, a center luminance value calculating circuit and an output circuit. In the operations of the image processing circuit, the receiving circuit receives image data. The reference value calculating circuit determines a first reference value and a second reference value corresponding to a plurality of pixels of the image data. The center luminance value calculating circuit refers to the first reference value and the second reference value to generate a center luminance value. The output circuit determines output luminance values of the plurality of pixel values according to the image data, the first reference value and the second reference value.

Abstract: A method for point cloud decoding includes receiving a bitstream. The method also includes decoding the bitstream into multiple frames that include pixels. A portion of the pixels are organized into patches and correspond to respective clusters of points of a 3D point cloud. The method further includes decoding, from the bitstream, an occupancy map frame. The occupancy map frame indicates the portion of the pixels included in the multiple frames that represent the points of the 3D point cloud. In addition, the method includes reconstructing the 3D point cloud using the multiple frames and the occupancy map frame. The method also includes determining whether to perform smoothing to the 3D point cloud based at least in part on properties of the multiple frames. Based on determining to perform the smoothing, the method includes performing the smoothing to the 3D point cloud.

Abstract: An electronic device is provided. The electronic device includes an input interface and a processor that performs noise removal processing on an image input through the input interface, obtains first information on a first texture block among a plurality of first pixel blocks included in the input image, obtains second information on a second texture block among a plurality of second pixel blocks included in the noise-removed image, obtains third information on maximum energy amount among energy amount of each of the plurality of second pixel blocks, and identifies whether the input image is an upscaled image based on the first, second and third information.

Abstract: A hybrid-distance adversarial patch generator can be trained to generate a hybrid adversarial patch effective at multiple distances. The hybrid patch can be inserted into multiple sample images, each depicting an object, to simulate inclusion of the hybrid patch at multiple distances. The multiple sample images can then be used to train an object detection model to detect the objects.

Abstract: A system and method for image reconstruction are provided. A first region of an object may be determined. The first region may correspond to a first voxel. A second region of the object may be determined. The second region may correspond to a second voxel. Scan data of the object may be acquired. A first regional image may be reconstructed based on the scan data. The reconstruction of the first regional image may include a forward projection on the first voxel and the second voxel and a back projection on the first voxel.

Abstract: The embodiments of the disclosure provide a method for transmitting reduced depth information and an electronic device. The method includes: obtaining a first depth map of an image frame, wherein the first depth map includes a plurality of first depth pixels corresponding to multiple image pixels in the image frame; downscaling the first depth map as a second depth map, wherein the second depth map includes multiple second depth pixels, and each second depth pixel has a depth value; adjusting the second depth map as a third depth map by scaling up the depth value of each second depth pixel based on a specific ratio; dividing the third depth map into multiple fourth depth maps; appending the fourth depth maps to the image frame; encoding the image frame appended with the fourth depth maps; and transmitting the encoded image frame to a client device.

Abstract: There is a need for more effective and efficient predictive data analysis solutions for processing genetic sequencing data. This need can be addressed by, for example, techniques for performing predictive data analysis based on genetic sequences that utilize at least one of cross-variant polygenic risk modeling using genetic risk profiles, cross-variant polygenic risk modeling using functional genetic risk profiles, per-condition polygenic clustering operations, cross-condition polygenic predictive inferences, and cross-condition polygenic diagnoses.

Abstract: A pigment detection method includes: extracting a first image from a to-be-detected RGB skin image, where the first image is used to represent a body reflection component in the RGB skin image, and the RGB skin image is photographed by a device having an RGB image photographing function; extracting a pigment from an R channel, a B channel, and a G channel of the first image based on a correspondence between a first spectral response curve of the pigment and a second spectral response curve of the device having the RGB image photographing function; and generating a pseudo-color image based on the extracted pigment, and displaying the pseudo-color image.