13results about How to "Accurate contour" patented technology

The invention relates to the field of image segmentation, in particular to a multi-modal nuclear medicine image segmentation method and system and a storage medium. Comprising the steps that image type classification is carried out on multi-modal medical examination images, focus area positioning is carried out, and CT (computed tomography) focus images and PET (positron emission tomography) focus images are obtained respectively; carrying out lesion geometric mutation form prediction on the CT fault lesion image, and carrying out multi-modal image fusion boundary correction on the CT fault lesion image and the PET lesion image to obtain multi-modal lesion image fusion boundary correction data; according to the multi-modal lesion image fusion boundary correction data, performing multi-modal lesion region segmentation combination optimization on the CT image and the PET image to obtain a lesion region segmentation combination optimization strategy; and carrying out image segmentation model construction on the lesion region segmentation combination optimization strategy based on a strategy gradient algorithm to obtain a lesion image segmentation model. According to the method, the image segmentation technology is optimized, so that the image segmentation technology is more perfect.

The utility model provides a steel brick composite flange brick for forming rolled glass, which is arranged on two sides of the top surface of a lip brick, a calender is arranged behind the lip brick, the calender comprises an upper calendering roller and a lower calendering roller, the steel brick composite flange brick comprises a flange brick body, a top end part a and a bottom end part b, cambered surfaces arranged at the upper part and the lower part of the tip part a are respectively matched with the upper calendaring roller and the lower calendaring roller, and a clamping part a clamped on a lip tile is arranged on the inner side of the cambered surface at the lower part of the tip part a; and at least the part between the upper calendering roller and the lower calendering roller is arranged in a profiling manner according to the outer contour of the tip part a, and the baffle structure is made of heat-resistant steel. By arranging the steel baffle structure which is profiled with the outer contour of the flange brick on the inner side of the tip end of the flange brick, the steel baffle structure provides an accurate contour, the flange brick body provides structural strength and heat preservation, the steel baffle structure is small in area, the heat dissipation area and the heat dissipation speed are greatly reduced, the production cost is reduced, and the edge clearing loss is greatly reduced; the problem that in the prior art, a steel flange brick is large in heat dissipation area and too fast in heat dissipation, so that the glass edge is crystallized, and edge loss is large is solved.

The invention relates to the technical field of image data processing, in particular to a special-shaped part production quality inspection method, device and system, and the method comprises the steps: obtaining three-dimensional point cloud data of a special-shaped part and a normal vector thereof, and extracting a key point and a corresponding normal vector thereof; clustering the key points based on similar features between the normal vectors and distance features between the key points; performing principal component analysis on the key points in each cluster to obtain a local shape feature value of each cluster; according to the distance distribution between the key points in each cluster, combining the local shape feature values to obtain an adjustment index of each key point; and adjusting the radius in the alpha shape algorithm to obtain the contour of the special-shaped part, and performing similarity analysis on the contour of the special-shaped part and the contour of the special-shaped part of the standard template to obtain a quality inspection result of the special-shaped part. According to the method, the radius in the alpha shape algorithm is optimized according to the characteristics of the special-shaped part, so that the effectiveness of boundary extraction is ensured, and the accuracy of a quality inspection result is improved.

The invention discloses a parking space editing method, device and equipment based on data synthesis and a medium, and the method comprises the steps: obtaining a parking space image and a label file thereof, and carrying out the semantic segmentation of the parking space image through a semantic segmentation model, and obtaining a parking space line mask; eliminating a parking space line in the parking space image based on the parking space line mask to obtain a target parking space image; modifying the line type or color of the parking space line mask to obtain a target parking space line mask; and performing data fusion on the target parking space line mask and the target parking space image to obtain a new parking space image, and modifying the tag file according to the attribute of the parking space line in the target parking space line mask. According to the method, the accurate parking space line contour can be obtained through the semantic segmentation technology, the parking space lines can be edited accurately, then the parking space lines are edited according to requirements to obtain different types of parking space lines, and the label of the original parking space image can be reused, so that the labeling cost and time for editing and generating new data are greatly reduced.

The invention provides an electronic device, a ceramic substrate and a preparation method thereof, and relates to the technical field of semiconductors. The ceramic substrate comprises a ceramic base body and a liquid channel layer, the liquid channel layer is arranged in the ceramic base body, the liquid channel layer and the ceramic base body are integrally sintered and formed, the liquid channel layer comprises a liquid channel, the liquid channel is used for containing a cooling medium, and the liquid channel extends in the ceramic base body along a three-dimensional space path. The preparation method of the ceramic substrate comprises the following steps: preparing slurry containing ceramic powder, preparing a sacrificial template matched with a target liquid channel in shape, pouring the slurry to coat the sacrificial template, filling a mold with the slurry to obtain a ceramic blank, removing the sacrificial template to form a hollow channel in the ceramic blank, and sintering to form the ceramic substrate with the liquid channel. The electronic device comprises the ceramic substrate, a component and a liquid cooling circulation system.

The invention belongs to the technical field of medical image segmentation, and provides a three-dimensional coronary artery segmentation method and system based on a CNN and a Transform, and the method comprises the steps: constructing a neural network model comprising an encoder and a decoder based on the CNN and the Transform, and enabling a bottleneck layer in the neural network model to comprise a multi-scale feature interaction module; inputting a three-dimensional coronary artery image to be segmented into the neural network model; wherein the features of the three-dimensional coronary artery image to be segmented are extracted in the encoder, and the features are integrated at a bottleneck layer through the multi-scale feature interaction module; and carrying out weighted fusion on the integrated features and sampling features output by each layer of the decoder in the decoder, and outputting a segmentation result. According to the invention, the defect of fragmentation of the segmentation result in the prior art is overcome, and a more accurate segmentation method is provided.

The invention relates to a disc brake for a motor vehicle and a motor vehicle with a disc brake, in particular a fixed caliper brake for a commercial vehicle (102). The invention relates to a disc brake for a motor vehicle (100), comprising a brake caliper (6) and two brake pads (8, 8 ') which are accommodated on the brake caliper (6) and can be moved relative to one another, and a brake carrier (10) for accommodating the brake caliper (6), the brake carrier (10) being designed to couple the brake caliper (6) to a rigid axle part (4) of the motor vehicle (100). The brake carrier (10) and the brake caliper (6) form a one-piece housing part (12, 12 ') which is designed to at least partially close the brake disc (2) which interacts with the brake linings (8, 8').

The invention relates to the technical field of image data processing, in particular to a resistor array infrared imaging control method and system based on transfer learning, and the method comprises the steps: obtaining an infrared image through a resistor array, and enabling each resistor unit in the resistor array to correspond to a range in the infrared image, each pixel point of the infrared image corresponds to one infrared radiation data value, and a thermal radiation area and a thermal crosstalk area of each resistor unit are determined according to the number of the pixel points in the corresponding range of the resistor unit on the infrared image, the infrared radiation data corresponding to the pixel points and the distance between central points of the corresponding ranges of the adjacent resistor units; and according to the area of the thermal crosstalk region, determining the influence degree of thermal crosstalk on the infrared image, and according to the influence degree, controlling the infrared imaging of the resistor array. According to the method, analysis is carried out based on infrared image data generated by transfer learning, a high-quality resistor array infrared image is generated by controlling the magnitude of current, and the influence of thermal crosstalk on the infrared image is reduced.

The application provides a depth image acquisition method, an electronic device and a storage medium. The method comprises the following steps: acquiring a first image and a second image; processing the first image to obtain a predicted depth map, and calculating a first error value by using a first preset loss function; determining a first transformation matrix between the first image and the second image; performing instance segmentation and mask processing on the first image to obtain a first mask image and a second mask image; processing the first transformation matrix according to the first mask image and the second mask image to obtain a target transformation matrix; generating a third image based on the target transformation matrix; calculating a second error value by using a second preset loss function; adjusting a depth learning network model according to the first error value and the second error value to obtain a target depth learning network model; and inputting a to-be-detected image into the target depth learning network model to obtain a depth image corresponding to the to-be-detected image. The application can obtain a more accurate depth image.

The application relates to the technical field of automobile foot mat processing equipment, and discloses a die forming device for an automobile foot mat, which comprises a bottom box, a top plate arranged above the bottom box, a concave die seat arranged at the top of the bottom box, a forming mechanism arranged on the top plate, a same back plate fixedly installed at the rear side of the top plate and the rear side of the bottom box, a piston cylinder fixedly installed on the inner wall of the bottom box bottom, and a piston mechanism arranged in the piston cylinder. The application has the following advantages and effects: the piston mechanism and the ejection mechanism are arranged, the piston plate can be driven to move upwards when the hydraulic cylinder output shaft is contracted after die forming, high-pressure gas can enter the hollow seat through the inflation pipe, and the high-pressure gas can be blown upwards through the air pipe, the generated air pressure can drive the supporting plate to move upwards against the spring elastic force, the formed foot mat can be ejected from the concave die seat, the foot mat can be separated from the die cavity, and the purpose of manual or mechanical arm material taking is achieved.

This invention discloses a visual inspection device for the end face of a tablet pressing mold, comprising a cabinet and a workpiece. The cabinet houses a worktable with a clamping mechanism for limiting and clamping the workpiece. A first visual inspection mechanism is mounted on the worktable, and a second visual inspection mechanism is located on the upper part of the worktable near the clamping mechanism. Compared to existing technologies, this device, with its adjustable clamping mechanism and two visual inspection mechanisms, effectively improves inspection efficiency through their combined operation. The clamping mechanism rotates the workpiece under the drive of a stepper motor. The first visual inspection mechanism integrates an end face inspection camera and a laser profilometer, simultaneously measuring end face parameters and scanning surface defects. The second visual inspection mechanism is equipped with multiple axial inspection cameras, enabling precise modeling of the overall contour of the workpiece, thus avoiding efficiency losses caused by multiple clamping operations.

The present application provides a method for detecting a machining surface of an engine and a computer-readable storage medium. The method includes: determining a plurality of continuous edge search segments of an image of a machining surface of an engine; determining a plurality of regions of interest of the machining surface image according to the plurality of edge search segments, and performing edge detection on the plurality of regions of interest to obtain an initial contour of the machining surface of the engine, wherein the regions of interest correspond to the edge search segments one by one, the edge search segments are located in the regions of interest, and the initial contour includes a plurality of initial coordinate points that are adjacent in sequence; updating the initial contour according to a preset edge distance to obtain a final contour, wherein the final contour includes a plurality of final coordinate points that are adjacent in sequence, the second segment is perpendicular to the third segment, the length of the second segment is the edge distance, the second segment is a segment formed by the initial coordinate point and the corresponding final coordinate point, and the third segment is a segment formed by the initial coordinate point and another adjacent initial coordinate point.