Abstract: An image processing method and a data set collection method is disclosed. The image processing method includes: making a high-resolution high-definition image interact with a calibrated blur kernel to obtain a blurred image; injecting calibrated noise to form a first noise image; degrading, by using a downsampling method, the first noise image into a low-resolution image; performing lossy compression on the low-resolution image to obtain a compressed image; adding into the compressed image random noise points that simulate a transmission process, to form a second noise image; and performing lossy compression on the second noise image to obtain a low-definition image, where the high-definition image is an image that is formed by means of photographing an organ in a body using a capsule endoscope, and the calibrated blur kernel and calibrated noise are results that are obtained by means of fitting a photographing process of the high-definition image.

Abstract: An apparatus includes a transceiver configured to receive, over a wireless communication channel, at least one controlled-noise signal. The apparatus also includes a processor, operatively coupled to the transceiver. The processor is configured to train, based on the at least one controlled-noise signal, a noise prediction model for the wireless communication channel, and generate, based on the trained noise prediction model, a noise prediction for the wireless communication channel. The processor is also configured to determine, based on the received at least one controlled-noise signal and the noise prediction, a score function for the wireless communication channel.

Abstract: Disclosed in the present invention are a dynamic maximal clique enumeration device and method based on an FPGA with an HBM, the method including: the HBM stores a dynamic edge flow, a complete graph adjacency matrix, and candidate cliques; a matrix computing unit updates the complete graph adjacency matrix based on the dynamic edge flow, transmits the updated complete graph adjacency matrix to the HBM for storage, and determines header nodes, of which the corresponding candidate clique needs to be updated; a sequence computing unit constructs, according to the updated complete graph adjacency matrix and each header node to be updated, the sorted data set for reconstructing candidate cliques by data block sequencing; and an update computing unit executes, in parallel, an update task of the candidate clique corresponding to each header node to be updated based on the sorted data set, transmits the updated candidate cliques to the HBM for storage, and transmits the updated candidate cliques to the PC host to ext

Abstract: An optical display apparatus is disclosed. A coupling apparatus couples light output by an image generation apparatus into an optical expansion waveguide. The optical expansion waveguide at least comprises a first reflective surface array and a second reflective surface array. Each reflective surface array is used for expanding, in an one-dimensional direction, light propagating through the optical expansion waveguide, so that the light entering into the optical expansion waveguide is expanded in a two-dimensional direction and is emitted outside of the optical expansion waveguide to form an image, wherein at least one of a first reflective surface of the first reflective surface array and a second reflective surface of the second reflective surface array is at least partially a curved surface, allowing the angle of emergence of the light to be changed by means of the first reflective surface or/and the second reflective surface.

Abstract: The present invention provides an interface-detection device and method. An internal interface of the device is connected to a protected device, an external interface is connected to an external device, and the external interface is in one-to-one correspondence connection to the internal interface by means of a switch module. At least two of the external interfaces are used for connecting to a predetermined data transmission type of external device, and each internal interface is used for one-to-one connection to all of the external interfaces of the protected device. The present invention detects the type of external device connected to any external interface of an externally-connected protection device, and, according to the detection result, instructs a switch device to connect or disconnect the external interface to/from a corresponding internal interface, and perform an alert operation.

Abstract: A method of performing a camera to ground alignment for a camera system on a vehicle. The method includes determining if enabling conditions have occurred and estimating a location of a vanishing point in a source image. Ground lines are selected based on a source image. A lane line detection is performed based on clustering of the ground lines to determine lane lines in the source image. At least one of pitch, yaw, or roll of the vehicle are estimated from the source image. A cost function based on estimates of pitch, yaw, and roll is minimized to obtain an optimal pitch value, an optimal yaw value, an optimal roll value, and lane lines from the source image. A sliding window-based refinement is performed on source images. Alignment results are broadcast to a downstream application or it is determined if the camera system on the vehicle is misaligned.

Abstract: Provided are a method and an apparatus for implementing batch extraction of human anatomical feature parameters. The method includes: obtaining a to-be-measured sample; generating an average model; calculating a point correspondence of a spatial location formed by the average model and each to-be-measured sample; measuring each to-be-measured sample on the average model; and outputting feature parameter data of each to-be-measured sample in batch. Calculation time is reduced by combining feature point extraction with affine transformation, and corresponding points are calculated through a non-rigid registration method to achieve batch and quick measurement.

Abstract: Provided are a method and apparatus for evaluating anatomical adaptation of an implantable orthopedic medical device. The method includes: obtaining an anatomical adaptation region of a to-be-evaluated bone with the implantable orthopedic medical device; determining a baseline of the anatomical adaptation region and a baseline of the implantable orthopedic medical device for position registration, and conducting the position registration for the anatomical adaptation region and the implantable orthopedic medical device based on the baselines; and evaluating adhesion of the to-be-evaluated bone to the implantable orthopedic medical device by calculating a distance from each point on an inner curved surface of the implantable orthopedic medical device to the anatomical adaptation region.

Abstract: Provided are a T-shaped reamed pile, a composite foundation, and a construction method for the composite foundation, relating to the technical field of civil engineering. The composite foundation is composed of a T-shaped reamed pile, soil between piles, and a cushion layer. The T-shaped reamed pile is composed of a core pile, a reamed body surrounding the outside of the core pile. A depth of the core pile is not lower than that of the reamed body, the reamed body includes an upper reamed body, and a lower reamed body, and a diameter of the upper reamed body is larger than that of the lower reamed body. The core pile is a precast pile, the reamed body is a cast-in-place reamed body, and the cushion layer is a sand-gravel cushion layer, or a reinforced cement-soil cushion layer.

Abstract: A refrigerator, comprising: a refrigerator body having an open front side to define a first compartment; a primary door mounted to the refrigerator body and used for opening/closing the first compartment, the primary door defining a second compartment having an open front side; and a secondary door mounted to the primary door and used for opening/closing the second compartment, the secondary door comprising a state adjustable door panel, and the state adjustable door panel being configured to controllably change transparency such that the visibility of an internal structure of the second compartment is adjustable. According to the refrigerator of the present invention, the frequency of opening the secondary door can be effectively reduced, the product high-tech element is added, and the user experience is improved.

Abstract: The present application is applicable to the field of semiconductor technology and provides a method for packaging stacking a flip chip, which includes: placing a filling template on a substrate, the filling template being provided with a through hole of a preset pattern; filling a filling material into the through hole of the filling template, and after the filling material being formed on the substrate, removing the filling template; placing a chip with solder balls on the substrate formed with the filling material, such that at least a portion of the solder balls being covered by the filling material; and connecting the chip to the substrate through the solder balls, and curing the filling material with air gaps formed between the at least a portion of the solder balls covered with the filling material.

Abstract: An anti-inflammatory drug compound, and a preparation method therefor and the use thereof. The structural formula of the compound is A-Y—B, wherein A is a group after dehydrogenation of an amine compound having JAK inhibitory activity, Y is a direct connection or —(CH2)-O— or, and B is a group formed by means of dehydroxylation of a carboxy-containing carboxylic acid compound B1, or a group formed by means of dehydrogenation of a hydroxy-containing compound B2. The compound has the special effects of having a strong transdermal property, controlled drug release, high efficacy, etc.

Abstract: The present invention provides a refrigerator door with a replaceable door panel, comprising: a foam door body, and a front panel detachably connected to the foam door body; a first limiting member and a second limiting member are disposed on a front side of the foam door body; a frame is sandwiched between the front panel and the foam door body, the frame is attached to a rear side of the front panel, a hook is disposed on the frame, and the hook has an insertion plate portion; wherein when the front panel is assembled on the front side of the foam door body, the first limiting member, in cooperation with the insertion plate portion, limits a displacement of the front panel in a first direction; the second limiting member, in cooperation of the frame, limits a displacement of the front panel in a second direction.

Abstract: An electronic device may include a lenticular display. The lenticular display may have a lenticular lens film formed over an array of pixels. A plurality of lenticular lenses may extend across the length of the display. The lenticular lenses may be configured to enable stereoscopic viewing of the display such that a viewer perceives three-dimensional images. Crosstalk between viewing zones and disparity between images received from different viewing zones may result in disparity-caused shifts in images perceived by viewer of the lenticular display. To mitigate these disparity-caused shifts, compensation circuitry may be included in the display pipeline circuitry. The compensation circuitry may include stored disparity-caused shift calibration information that is used for the compensation. The stored disparity-caused shift calibration information may be a polynomial function that outputs a magnitude of disparity-caused shift for a given pixel location.

Abstract: A method for manufacturing a solar cell, a method for manufacturing a monocrystalline silicon wafer and a photovoltaic module. The method for manufacturing a monocrystalline silicon wafer includes: providing a monocrystalline silicon rod; squaring the monocrystalline silicon rod to form a quasi-square silicon rod with quasi-square cross-section having an arc, a length of the arc being not less than 15 mm; slicing the quasi-square silicon rod to form at least one quasi-square silicon wafer having the arc. The method for manufacturing at least one solar cell includes: using the method described above to obtain a quasi-square silicon wafer having an arc; forming a first solar cell by processing the quasi-square silicon wafer; scribing the first solar cell to obtain a square-shaped sub-solar cell and at least one strip-shaped sub-solar cell. The above methods improve the utilization rate of the monocrystalline silicon rod and reduce production cost.

Abstract: A virtual reality display method and apparatus, and a storage medium are provided. The virtual reality display method includes: rendering display data of a plurality of screens in sequence by utilizing one thread of at least two threads, and after rendering the display data of the plurality of screens, correcting the rendered display data by using the at least two threads.

Abstract: Provided are a method and system for controlling a Raman fiber amplifier. The method comprises: according to a target gain and a tilt, calculating an expected output power of a pump by using a feedforward formula, and obtaining an actual output power of the pump through detection (201); locking the actual output power of the pump to the expected output power through first-stage feedback control (202); according to the target gain and the tilt, calculating an expected ASE power of the pump by using an ASE formula, and obtaining an actual out-of-band ASE power of the pump through detection (203); if the out-of-band ASE is not locked, determining gain compensation and tilt compensation of the pump through second-stage feedback control, and feeding the compensation back to the feedforward formula and the ASE formula for recalculation (204); and repeatedly performing the first-stage feedback control and the second-stage feedback control until the gain and the tilt are locked (205).

Abstract: The audio playing method is provided by embodiments of the present disclosure, applied to on-board audio system (including multiple speaker modules arranged at different positions in a car cabin, the audio playing method includes: obtaining an audio signal to be played; decomposing the audio signal to obtain multiple audio source signals of different categories, where each of the multiple sound source signals includes multiple sub sound source signals of a same category; establishing a route between all sub sound source signals and the multiple speaker modules according to a preset distribution principle indicating rules for playing different sub sound source signals by different speaker modules; and playing each of the multiple sub sound source signals by a corresponding speaker module according to the route. The present disclosure provides more accurate and immersive car multi-channel surround sound playback, which can effectively improve overall acoustic experience of the on-board audio system.

Abstract: Provided by the present disclosure is an ultrasonic phased array-based in-situ imaging method for melt flow in injection molding. The ultrasonic phased array is used for the detection of an injection molding process for the first time, and an effective dynamic monitoring imaging method for a melt front position is developed. A melt flow process in a mold cavity is dynamically monitored by collecting an FMC (Full matrix capture) dataset online. A mapping relationship between an incident angle and a target pixel point is established to rapidly determine time delay of each point in a measurement target region, and a melt bottom image is acquired using TFM (Total focusing method) imaging conditions, from which the melt front can be localized. The provided method is high in measurement accuracy, short in imaging time, and capable of effectively improving imaging efficiency of online measurement.

Abstract: A method for training a contrastive learning model, acquiring an optimization item, and pushing information includes: obtaining sample data including first category information and query information; predicting a correlation between the first category information and the query information using the contrastive learning model; establishing a loss function based on the prediction results, wherein the loss function is configured to characterize accuracy of the prediction results; optimizing the loss function based on a semantic relationship between the first category information and second category information, wherein the first category information and the second category information correspond to different semantic information in a category tree, and the semantic relationship is related to relative positions of the first category information and the second category information within the category tree; and training the contrastive learning model based on the optimized loss function.