228results about How to "Prevent disappearing" patented technology

The controller is programmed to cause a low-surface-tension liquid supply unit to supply a liquid film of a low-surface-tension liquid to a front surface of a substrate so as to form a liquid film of the low-surface-tension liquid. The controller is programmed to control the substrate rotating unit and the inert gas supply unit so that an inert gas is supplied toward the rotational center position while rotating the substrate, thereby forming an opening spreading from the rotational center position to be formed in the liquid film, and enlarging the opening in a direction away from the rotational center position, and to control the landing-position changing unit to change the landing position of the low-surface-tension liquid to at least two positions except the rotational center position in accordance with enlargement of the opening so that the landing position is placed outside the peripheral edge of the opening.

The invention provides a method for forming a solder bump. The method comprises the following steps of: forming a heat resistant metal layer and a metal wetting layer on a pad and a passivation layer of a chip in sequence; forming a photoresist on the metal wetting layer, wherein the photoresist is provided with an opening to expose the metal wetting layer above the pad of the chip; forming an adhesion layer and a barrier layer on the metal wetting layer in the opening in sequence; forming solder paste on the barrier layer; removing the photoresist; etching the heat resistant metal layer and the metal wetting layer on the passivation layer until the passivation layer is exposed; and ensuring the solder paste to reflow to form a pillar bump. The method has the effect of improving the electrical property and reliability of the solder bump and is suitable for chip scale package with the requirements of fine pad pitch and multiple output functions.

The embodiment of the invention discloses a fatigue detecting method based on a deep space-time network, and relates to the technical field of machine learning. The method comprises the steps that driving environments are simulated, wherein the driving environments comprise driving environments under different weathers and road conditions; sample data is collected and classified under the drivingenvironments, wherein the sample data comprises a first data set and a second data set; a deep space-time network model is constructed, and a feature extractor is generated to extract eye features inthe sample data; the eye features of the first data set are subjected to feature compression by using an automatic codec and input into a recurrent neural network RNN; the second data set is used fortraining the recurrent neural network RNN to obtain a trained fatigue detecting model; the trained fatigue detecting model is used for real-time detection, a detection result is output, and feedback is performed through a warning module. The fatigue detecting method based on the deep space-time network can solve the problems that the fatigue detecting method in the prior art cannot accurately classify the fatigue degrees and is inconvenient to operate.

The invention discloses a dangerous working area accident automatic detection alarm method based on deep learning, and the method comprises the steps: obtaining original video data, carrying out the preprocessing, and converting a video into an input training set acceptable for a deep learning network; learning a feature mode in a training video through a convolutional space-time automatic encoderdecoder, and performing training optimization by using the training set to obtain a workshop accident detection model; and acquiring a real-time monitoring video to be detected, detecting the reconstruction error of each frame of monitoring video image by adopting the workshop accident detection model, and if the local minimum reconstruction error of a plurality of continuous real-time monitoringimages is greater than a threshold value, sending corresponding alarm information and corresponding monitoring position information to a workshop administrator terminal. On the basis of analysis of alarge number of videos, video special learning of a normal scene is carried out, a fully trained detection model is obtained, abnormal accidents in a workshop can be rapidly and accurately detected,and accident detection can be carried out in any workshop scene.

Generative adversarial networks (GAN) display a great development prospect in image generation and semi-supervised learning, and have been developed to Triple-GAN. However, there are two problems requiring solving based on the classification method based on Triple-GAN, wherein the two problems comprise a problem that KL-based divergence distribution construction is easy to generate gradient disappearance to cause unstable training and a problem that the Triple-GAN employs a manual mode to perform mark of tags for samples to cause too large workload of manual marking and non-uniform marking. Based on this conditions, RandomForests are employed to perform classification of real samples to perform automatic marking of tags of blade nodes, and the ideal of the least square generative adversarial networks (LSGAN) are employed to construct a loss function to avoid gradient disappearance.

An image-level detecting section detects luminance levels in detection areas specified by a video signal processing control section. The video signal processing control section analyzes the detection results to determine the position and range of a non-picture area, and then calculates an optimum vertical magnification which allows only a video signal area to be displayed in the same size as the screen size of an output monitor. A video signal processing circuit clips only the video signal area based on a control signal provided from the video signal processing control section and magnifies the video signal area, thereby displaying the video signal area that should be viewed without causing any disappearance thereof.

A columnar bump packaging process comprises the following steps: forming lower metal bump layers on a welding pad and a passivation layer of a chip; forming photoresist on the lower metal bump layers, wherein the photoresist is provided with an opening so as to expose the lower metal bump layer above the welding pad of the chip; forming a copper column on the lower metal bump layer in the opening; forming a welding bump on the copper column; removing the photoresist; etching the lower metal bump layer on the passivation layer till the passivation layer is exposed; forming an oxide layer on the surface of the exposed copper column; and removing oxide on the surface of the welding bump, and refluxing the welding bump. The columnar bump packaging process of the invention improves the electrical performance and the reliability of the welding bump and is applicable to chip-level packaging of the welding pad with a dense spacing and multiple output functions.

The invention relates to a cylindrical bump packaging structure. The cylindrical bump packaging structure comprises a chip, an under bump metallization layer, copper columns, oxide layers and solder bumps, wherein bonding pads and a passivation layer are arranged on the upper surface of the chip, the passivation layer is covered on the upper surface outside openings of the bonding pads of the chip, the under bump metallization layer is arranged on the bonding pads, the copper columns are arranged on the under bump metallization layer, the oxide layers are wrapped on the side surfaces of the copper columns, and the solder bumps are arranged above the copper columns. According to the cylindrical bump packaging structure disclosed by the invention, the electrical performances and the reliability of the solder bumps can be improved, and the cylindrical bump packaging structure is suitable for chip level packaging with close intervals of the bonding pads and multiple output functions.

The invention discloses a temperature sensitive polymer material with biodegradability and biocompatibility. A preparation method of the material comprises the following steps of: firstly synthesizing monomers with specific structures by using the basic raw materials including lysine, tartaric acid, and malic acid, and carrying out polymerization reaction on the monomers with different ingredients to prepare the heterochain type polymer material. The polymer not only has good temperature sensitivity, but also has good biodegradability and biocompatibility. The preparation method disclosed by the invention is simple and easy to implement and has good industrial prospect.

The invention discloses a method and a system for upgrading a CPLD based on a Whitley platform. The method comprises the following steps: a CPLD code is burned into a flash memory of the CPLD througha background mode; Through the JTAG interface of CPLD, BMC controls the pin of CPLD in the locked state,the CPLD code in the CPLD flash memory is loaded into the static random access memory of the CPLD, By controlling the level of the Hitless_EN signal, the BMC releases the pin state of the CPLD. The system includes four parts: a burning module, a first locking module, a loading module and a releasing module. The method and the system in the present application can avoid the out-of-control state of the CPLD pin lock, thereby avoiding the unstable state of the performance that may be caused inthe CPLD upgrading process, greatly improving the stability and reliability of the CPLD upgrading, and truly realizing seamless burning and recording.

The invention provides a webpage refreshing method, a webpage system and electronic equipment, the webpage comprises at least one sub-frame, the sub-frame is used for displaying one page in at least one target page, and the method is characterized by comprising the following steps: storing a page identifier of the target page displayed by the sub-frame as a page identifier record; when a page is loaded, obtaining the page identifier; and setting a target page to be displayed by the sub-frame according to the acquired page identifier. By the adoption of the webpage refreshing method, system andequipment, the sub-frame page content before refreshing can be reserved after the user refreshes the webpage, and the user experience is greatly improved.

The invention discloses a cutting propagation method for an ancient platycladus orientalis. The method includes the steps: sequentially using potassium permanganate, sodium lignite and hormone for treating cuttings of the ancient platycladus orientalis to obtain treated cuttings; subjecting the treated cuttings to cuttage and cultivation in a matrix to realize cutting propagation of the ancient platycladus orientalis. According to experimental results, the cutting propagation method has the advantages that survival rate in clonal propagation of the ancient platycladus orientalis is increased, rooting time of the ancient platycladus orientalis is shortened, and references are provided for research on clonal propagation of other ancient trees.

A relay server is capable of communicating with other relay servers through an external server arranged on a Wide Area Network. The relay server stores information about a registered terminal, information about the relay servers forming a relay group by permitting connection among each other, and information about a resource and a client terminal sharing the resource. When an operation instruction regarding the resource in another terminal is received, the relay server relays the operation instruction from the registered terminal to another relay server. However, when key information is stored in association with the resource, the relay server relays the operation instruction to the other relay server only when a correct key has been entered.

The invention discloses an unsupervised remote sensing image semantic segmentation method based on super-resolution and domain self-adaption, and belongs to the technical field of remote sensing image semantic segmentation methods. The technical problem to be solved is to provide the improvement of the unsupervised remote sensing image semantic segmentation method based on super-resolution and domain self-adaption. According to the technical scheme, the method comprises the following steps: obtaining a source domain low-resolution remote sensing image data set and a target domain high-resolution remote sensing image data set, and dividing the obtained target domain image data set into a training image and a test image according to a set proportion; building a remote sensing image semantic segmentation network and a super-resolution network; carrying out network pre-training and parameter optimization on the built super-resolution network; training a remote sensing image semantic segmentation network; inputting the preprocessed test set data into the trained remote sensing image semantic segmentation network, and outputting an accurate segmentation result of the remote sensing image; the invention is applied to remote sensing image processing.

A battery power detecting system is disclosed applicable to an electronic device having a volatile random access memory (VRAM) for detecting the battery power of a backup battery of the volatile random access memory. The battery power detecting system includes a detection module for detecting the battery power of the backup battery and for generating a detection signal; a reception module for receiving the detection signal; and a determination module for reading a state value of the detection signal received by the reception module, determining the battery power of the backup battery according to the state value, and generating alarm messages to notify a user that the backup battery is lacking in battery power when the backup battery is determined to be have insufficient battery power, so as to prevent data stored in the VRAM from being lost due to the lack of the battery power of the backup battery.

The invention discloses a dam deformation prediction system and method based on a convolutional neural network. According to the convolutional neural network for dam deformation prediction, the network parallel computing capacity is fully used, two-dimensional data is input into multiple paths in parallel, multiple convolutional layers and pooling layers are alternately cascaded in each path, andin order to obtain more characteristics, different convolutional kernel sizes can be set; a simple path only cascading a convolution layer and a pooling layer is added to extract features different from other paths, and meanwhile, the features extracted by the paths are not necessarily complementary, so that additive operation is used in the system when output results of the paths are combined. According to the method, the automation level of dam deformation prediction is improved, and the efficiency and speed of deformation prediction are improved.

The invention relates to a panel for assembly and a machining technology thereof and a dryer. The machining technology comprises the steps that S1, plate blanking is carried out; S2, laser cutting iscarried out, specifically, laser cutting is carried out on a plate according to the boundary contour of a developing drawing of a finished panel; S3, slotting is carried out, specifically, a pluralityof bent lines are drawn on the plate, and slotting is carried out along the bent lines; S4, bending is carried out, specifically, bending treatment is carried out along slotted lines, one opposite sides are bent first, and then the other opposite sides are bend; and S5, welding and polishing are carried out, specifically, butt joints of the edges are subjected to welding and polishing treatment.According to the machining technology, five visible surfaces of the panel have no side seams which need to be welded, all positions which need to be welded are arranged on the inside side surfaces, namely, the invisible surfaces, line disappearance and chromatic aberration of the visible surfaces after welding and polishing are avoided, and an integral attractive effect is achieved. Meanwhile, themachining technology of the panel can effectively simplify the operation steps of bending and effectively ensure the overall flatness and attractive degree of bent parts.

The invention belongs to the technical field of casting, and relates to a technological method of using internal chill expendable pattern negative pressure casting for producing a dispersed hot spot shrinkage porosity qualified casting without a shrinkage cavity inside. The method includes the steps of selecting an internal chill material, determining internal chill quality, manufacturing the internal chill, cleaning the internal chill, manufacturing internal chill formed foam model, conducting molding, performing casting and obtaining the casting after solidification. The same material as the casting or a commercially available nickel plated wire structure is adopted in the internal chill, so that the internal chill the same as the casting in shape is manufactured conveniently. The straight rod and winding combined structural form is adopted in the internal chill, the straight rod structure has fixing and supporting functions and can prevent that the internal chill collapses an expendable pattern cavity due to excessive mass. The internal chill is simple in structure, low in cost, easy to operate, high in reliability and applicability and suitable for the casting provided with dispersed hot spots and formed by an expendable pattern.