50results about How to "No reduction in quantity" patented technology

The invention discloses an image brightness adjusting method, an image brightness adjusting device, computer equipment and a storage medium. The method comprises the steps of: on the basis of the brightness value of each pixel point in an image to be processed, obtaining an original brightness histogram of the image to be processed; on the basis of the original brightness histogram, obtaining thedark boundary brightness value and the bright boundary brightness value of the original brightness histogram; on the basis of the dark boundary brightness value and the bright boundary brightness value of the original brightness histogram, adjusting the brightness of the original brightness histogram, so that a target brightness histogram is obtained; fitting the target brightness histogram, so that a smooth curve is obtained; and, on the basis of the smooth curve, adjusting the brightness value of each pixel point in the image to be processed, so that a balanced image is obtained. Thereby, the brightness of the image to be processed can be adjusted rapidly.

The invention discloses a management method for positioning archival data to books. The management method comprises the steps as follows: (1), a plurality of rotatable hanging hoppers are mounted in an archive cabinet, and after information of archival data books is input into a terminal control system of the archive cabinet for storage, the archival data books are placed in the hanging hoppers in sequence; (2) the information of one archival data book is input into the terminal control system, one hopper stored with information of the archival data book is operated to an access window, a laser indicator light is switched on, and laser irradiates on the archival data book or the position where the archival data book is placed; (3), after the archival data book is stored, the single operation is completed; and (4), the step (2) and the step (3) are repeated, and the next archival data book can be positioned and stored. With the adoption of the management method for positioning the archival data to the books, the work efficiency of archive management staff can be improved, and the archival data can be searched accurately and positioned to books. The invention further relates to a device for realizing the management method for positioning archival data to the books.

The invention discloses a light-emitting diode (LED) linear current control circuit. The LED linear current control circuit comprises an LED lamp string, an input voltage sampling network, a comparison and control circuit and a multi-path selection switch, wherein the LED lamp string is formed by connecting at least one LED in series; the input voltage sampling network is used for sampling the input voltage of the LEDs; the comparison and control circuit compares the input voltage of each LED with a reference voltage to obtain a comparison result; and the multi-path selection switch is respectively connected to the LEDs and changes conduction impedance according to the comparison result so as to control the conduction of a part of the LEDs or all the LEDs. The invention also discloses an LED linear circuit.

The embodiment of the invention discloses a camera assembly, and is characterized in that the camera assembly comprises a housing, a reflection assembly, a driving assembly, and an imaging assembly; The shell is provided with a first light-transmitting assembly and a second light-transmitting assembly which are opposite in orientation and are aligned in the center. The driving assembly drives thereflecting assembly to rotate around a rotating shaft, so that the reflecting assembly is switched between a first position and a second position; The reflection assembly is located at a first position, and the reflection surface of the reflection assembly reflects the first ambient light passing through the first light transmission assembly to the imaging assembly; The reflection assembly is located at a second position, and the reflection surface of the reflection assembly reflects the second ambient light passing through the second light transmission assembly to the imaging assembly. The embodiment of the invention further discloses electronic equipment comprising the camera assembly and a control method of the camera assembly.

The invention provides a compression method and system for a neural network model, electronic equipment and a readable medium. The method comprises the steps: presetting a pruning strategy of each network layer, and the pruning strategy comprises a pruning rate, a pruning threshold and a filter weight value; in the current compression period, selecting a pruning filter to be pruned from each network layer according to a pruning strategy, and setting the value of each pruning filter to be zero; setting the value of the pruning filter to be a non-zero value through fine adjustment so as to update the neural network model and obtain a current neural network model; and determining whether the pruning strategy of the next compression period needs to be adjusted or not according to the current meta-attribute value output by the current neural network model. Allowing the pruned filter to update without reducing the number of feature maps of the network; an extra model fine adjustment stage isnot needed any more, and the network training time is shortened; according to the parameter statistical distribution of the current model, the pruning strategy most suitable for the current model parameters is selected, and the model training effect is improved.

The invention discloses a compression method and device of a deep neural network model, a terminal and a storage medium. The method comprises the following steps: acquiring a trained deep neural network model; traversing all layers of a deep neural network to obtain parameters of a current traversed layer; and quantifying the parameters of the current traversed layer based on the density and extension range of the parameters until all the layers of the deep neural network are traversed to obtain a quantified deep neural network model. Through the scheme, the accuracy and validity of the deep neural network model can be improved during compression of the deep neural network model.

The invention relates to a radiating device for an LED lamp and a radiator component. The radiator component comprises an arc-shaped plate. The outer arc surface of the arc-shaped plate extends outwards to form a plurality of spaced radiating fins, a connecting groove is arranged at one longitudinal end face of the arc-shaped plate along the longitudinal direction, a connecting convex strip is arranged at the other longitudinal end face of the arc-shaped plate along the longitudinal direction, and at least one heat-conduction pipe installing hole is longitudinally arranged at the upper edge of the arc-shaped plate. The bottom parts of the radiating fins are thick, the top parts are thin, and the convex strips are distributed on the surfaces of the radiating fins. The radiating device comprises a radiator, a heat-conduction piece and at least one heat super-conduction pipe. The radiator is composed of at least one radiator component. One end of the heat super-conduction pipe is inserted into the heat-conduction pipe installing hole on the heat-conduction piece, and the other end is inserted into the heat-conduction pipe installing hole on the radiator component. As the radiating device adopts the modular design, a mould of the radiator is single and the development cost of the mould can be saved. With good radiating effect and light weight, the radiating device can greatly save aluminum materials and reduce the cost.

The invention discloses a non-overshoot LED linear constant current drive circuit. The circuit comprises an input rectifier bridge, an input voltage detection network, a power supply network, an LED constant current drive circuit core control unit whose input end is respectively connected with the input voltage detection network and the power supply network, an LED control switch whose input end is respectively connected with the output end of the LED constant current drive circuit core control unit. One end of the input voltage detection network is connected with the input rectifier bridge, and the other end is grounded. The power supply network is formed by connecting a resistor and a capacitor in series, wherein one end of the power supply network is connected with the input rectifier bridge, and the other end is grounded. The output end of the LED control switch is connected with an LED lamp string. According to the invention, various disadvantages of a common LED linear drive circuit can be eliminated; the LED use efficiency can be effectively improved; and the non-overshoot LED linear constant current drive circuit has the advantages of low cost, high power factor, high engineering applicability and the like.

The embodiment of the invention discloses a data query method and a network communication system, and relates to the technical field of communications. The method disclosed by the invention is capable of providing the service address configuration for a large-capacity data plan service, and guaranteeing that the user number of accessing the data plan service cannot be reduced when the system can support the page view. The method disclosed by the embodiment of the invention comprises the following steps: a server receiving a service address of a target service sent by gateway equipment, and determining a service identify corresponding to the target service and a RG (rating group) corresponding to the target service according to the service address of the target service and a mapping table; the server sending the service identity corresponding to the target service and the RG corresponding to the target service to the gateway equipment. The method disclosed by the invention is suitable for the network communication system.

The invention discloses a point cloud smooth fairing filtering method based on a normal vector. The method comprises steps of firstly, removing outliers of an original point cloud; carrying out principal component analysis on the discrete point cloud set to deduce the normal vector of each point, and adjusting the normal vectors of all the points to be in the same direction; fitting a plane for each point by using a least square method through K neighbors of the point, obtaining a normal vector of the point through a plane model, and adjusting the normal vector to be in the same direction; correcting the normal vector calculated in the first step by utilizing the normal vector of the fitting plane; and finally, projecting the point to the fitting plane along the direction of the correctednormal vector, i.e., adjusting the position of the point to the intersection position of the corrected normal vector and the fitting plane so as to achieve an effect of carrying out smooth fairing processing on the point cloud. The method is easy to implement, the smooth fairing effect is obvious, the original point cloud is filtered to lay a foundation for triangularization of subsequent point cloud data and generation of DEM, and the method is suitable for fairing and smoothing the original point cloud data and is high in practical value.

The invention provides an industrial personal computer and an industrial personal computer communication protection method. The industrial personal computer comprises a main control module, a power module and a communication module. The main control module is connected with at least one peripheral device interface through the communication module. Each peripheral device interface is connected witha peripheral device. The peripheral devices communicate through the communication module. The main control module, the communication module and the peripheral devices are powered by the power module.During communication among the main control module, the communication module and the peripheral devices, when a data sender sends a data packet, serial number, source address and destination addressinformation is added to the data packet, and the check value of the data packet is calculated; the data sender packages and sends the information; and a receiver checks whether the source address andthe destination address of the data packet are correct or not after receiving the data packet. According to the industrial personal computer, an X86 architecture processor serves as a core, communication between the industrial personal computer and different peripheral function boards is achieved by expanding the number and types of high-speed buses output by a system board, and the purpose of function expansion is achieved.

The invention discloses an array substrate, an LCD (Liquid Crystal Display) panel and a display device. Under the situation that the numbers of data lines and grid lines are not reduced, two data lines are arranged in a gap between every two adjacent pixel structures through improvement of a data line structure, so that the area of a black matrix for covering the data lines in each pixel structure is reduced to a larger extent, and the aperture opening ratio is increased; meanwhile, through the improvement of the data line structure, the probability that disclination occurs can be further reduced, the utilization rate of liquid crystal is increased, and the transmissivity is further increased; in addition, the grid lines are utilized for shielding, so that the area of the black matrix for covering the grid lines is reduced, the aperture opening ratio and the transmissivity of the LCD panel are greatly increased through combined action of the grid lines and the black matrix, and the image quality of the LCD panel is further greatly increased.

The invention relates to an image processing network training method, computer equipment and a readable storage medium. The method comprises the following steps: acquiring a plurality of training sample images, wherein the plurality of training sample images are a plurality of groups of tomographic images; inputting the plurality of training sample images into an initial image processing network for image processing to obtain image processing results of the plurality of training sample images; and calculating the loss between the image processing results of the plurality of training sample images and the corresponding gold standards, calculating the difference value between the image processing results of the adjacent layers of training sample images in the same group of training sample images, and training the initial image processing network according to the loss and the difference value. According to the method, when the loss of the image processing result of the training sample image is calculated, the difference value between the image processing results of the adjacent layers of training sample images is considered, and meanwhile, the number of the training sample images is not reduced, so that the network precision obtained by training is greatly improved.

The invention aims to provide a semiconductor device manufacturing method, which is capable of reducing the width of a guard ring area on the premise that the guard rings are not reduced and the depth thereof are not decreased and has no complex and time-consuming processes for forming guard rings. A guard ring 124 is formed by selectively guiding a great amount of p type impurity from n type impurity diffusion area 120 on a guard ring area GR of a semiconductor basal body 110 by means of a second mask M1 and making the p type impurity perform diffusion inside the semiconductor basal body 110. The guard ring 124 has a structure which is located inside the n type impurity diffusion area 120 when viewed from a plane and is provided with a p type impurity diffusion area 122 in a position deeper than the lower surface of the n type impurity diffusion area 120 when viewed from the section.

The invention provides a suspension type urban overhead bus channel. The channel is characterized in that: piers are arranged in isolated grasslands between sidewalks on ground roads or driveways on the ground roads and non-motorized vehicle lanes; bridges are erected on the piers; overhead bus lanes are suspended below the bridges; overhead bus stops are arranged on the overhead bus lanes with regular intervals; the overhead bus stops are connected with the sidewalks on the ground roads through staircases or elevators; curve paths and traffic signal lamps are arranged in the longitudinal positions and the lateral positions of the intersection of the overhead bus lanes; local isolated balusters or continuous isolated balusters for the whole road section are arranged between the sidewalks on the road sections of gateways below the overhead bus stops and the non-motorized vehicle lanes or the driveways; the overhead bus lanes are communicated with the driveways on the ground through ramps, such that vehicles can pass through the overhead bus lanes through the driveways on the ground roads.

The invention discloses a quartz boat and a solar cell diffusion process. The quartz boat comprises at least three groove rods, wherein a plurality of tooth grooves are formed in each groove rod, are corresponding to one another and are used for placing silicon wafers, and the tooth distance between adjacent tooth grooves in the groove rod is increased in geometric progression by taking a first tooth distance as a reference. The solar cell diffusion process comprises a boat entering process, a constant-temperature process and a diffusion process. By the quartz board, the probability consistence that the silicon wafers contact with a diffusion source on different regions in a diffusion tube can be improved; moreover, by the solar cell diffusion process, the dosage of a nitrogen source can be reduced, the consistency of diffusion doping longitudinal concentration distribution in wafers and among wafers also can be improved, and the battery conversion efficiency consistency during mass production is further improved; and compared with a conventional diffusion process, the solar cell diffusion process has the advantages that the average efficiency of a battery piece which is prepared after employing the quartz boat and the diffusion process is improved by 0.05%, and the proportion of a low-efficiency battery piece with the battery conversion efficiency smaller than 18.0% is simultaneously reduced by 1.5%.

The invention discloses a sorting system, a sorting platform and a sorting unit thereof. The sorting unit comprises a plurality of partition ports formed in a platform body of the sorting platform, aswell as a plurality of slide grooves and a plurality of goods collection containers which are located below the platform body and as many as the partition ports, wherein one goods collection container is arranged below each partition port, and each partition port is butted with a slide groove inlet; and each slide groove is inclined, and configured to butt the partition port at each slide grooveinlet with the goods collection container arranged below the adjacent partition port of the partition port butted with the slide groove inlet, and the adjacent partition port is clockwise adjacent toor anticlockwise adjacent to the partition port butted with the slide groove inlet. The sorting system, the sorting platform and the sorting unit thereof are capable of greatly increasing the area ofa carrying channel in a site.

The invention relates to the technical field of drug detection, in particular to a sterile detection method of voriconazole for injection. Accordingt to the sterile detection method of voriconazole for injection, when a membrane filtration method is adopted to carry out sterile detection on voriconazole for injection, a sulfobutyl beta-cyclodextrin sodium solution with the concentration not higherthan 1% is adopted as a flushing fluid. According to the sterile detection method, the antibacterial activity of voriconazole can be remarkably reduced, the accuracy of the sterile detection result of voriconazole for injection is improved, meanwhile, the sample amount treated by each membrane during membrane filtration can be increased, and the detection cost of an enterprise is reduced.

The invention provides a low-power-consumption high-reliability semi-wrapped trench gate silicon carbide MOSFET device and a preparation method thereof. The low-power-consumption high-reliability semi-wrapped trench gate silicon carbide MOSFET device comprises an N-type substrate, an N-type epitaxial layer, a first P-body region, a first P + contact region, a first N + contact region, a second P-body region, a second P + contact region, a second N + contact region, a gate medium, a trench gate, an auxiliary trench gate, a source electrode and a drain electrode. Compared with a traditional half-wrapped trench gate silicon carbide MOSFET, the gate dielectric is protected by forming the second P-body region at the bottom of the trench, the reliability of the gate dielectric of the device is enhanced, a part of gate-drain capacitance is shielded, the switching loss of the device is reduced on the basis of not sacrificing the conduction capability of the original trench MOSFET, and when the device is short-circuited, the drain voltage is relatively large, so that the reliability of the device is improved. And at the moment, the JFET region between the second P-body region and the first P + contact region is pinched off, so that the saturation current of the device is reduced, and the short-circuit capability of the device is improved.

The invention provides a handheld electronic equipment charging device, which comprises a charging device body, wherein the body is in a cuboid shape; a power plug is arranged on the lower surface of the body; a power jack is arranged on the body; and a USB interface is arranged on the side surface of the body. The handheld electronic equipment charging device can improve the usage rate of the socket.

The invention relates to the field of recovery and storage of biological waste gases, more particularly to a storage device for greenhouse gases. The device comprises three portions, wherein a flatus storage tank mainly comprises an anus cover, a defecation sensor, a lower flatus pipe, a motor rotating shaft, an upper flatus pipe, a flatus tank, a flatus discharge pipe port and a shoulder belt; a flatus collector platform consists of a flatus tank connection pump pipe, a refrigerator, a pressurizer and a flatus collection platform; a compression liquid state flatus tank comprises a platform connection pump pipe, a pressurizer and a compression liquid state flatus tank. The shoulder belt is connected with the flatus tank and fixed on the back of a cow or sheep. The anus of the cow or sheep is connected with the anus cover, and the anus cover is connected with the lower flatus pipe. Furthermore, the defecation sensor is connected with the motor rotating shaft via the inner pipe of the lower flatus pipe. The motor rotating shaft is fixed on the upper flatus pipe and is connected with the lower flatus pipe. The upper flatus pipe is connected with the flatus tank. The flatus tank is also connected with the flatus discharge pipe port. The flatus discharge pipe port is connected with the flatus tank connection pump pipe and the flatus collector platform. The flatus connector platform equipped with the refrigerator and the pressurizer is connected with the compression liquid state flatus tank equipped with the pressurizer via the platform connection pump pipe. The invention has the advantage that: the device can efficiently collect and treat flatus (greenhouse gas) of cows and sheep, thereby preventing environmental damage caused by flatus that mainly contains ammonia gas (major component of acid rain) and methane and carbon dioxide (major components of greenhouse gas). Therefore, the device can protect the environment without affecting cow and sheep breeding.

The invention discloses a camera mechanism and electronic equipment. The camera mechanism comprises a shell, a light path conversion assembly and an imaging assembly. A first light-transmitting area and a second light-transmitting area are arranged on the two sides of the shell. The light path conversion assembly comprises a first transflective mirror, a second transflective mirror, a first reflecting surface and a second reflecting surface. When light enters the first light-transmitting area, part of the light is reflected by the first reflecting surface and then penetrates through the firsttransflective mirror to be emitted to the imaging assembly, and part of the light penetrates through the first light-transmitting mirror and then is reflected by the second light-transmitting mirror to the imaging assembly. When light enters the second light-transmitting area, part of the light is reflected from the second reflecting surface and then penetrates through the second transflective mirror to be emitted to the imaging assembly, and part of the light penetrates through the second light-transmitting mirror and then is reflected to the imaging assembly through the first light-transmitting mirror. The camera mechanism can effectively control the cost and the occupied space, the image photographing quality can be easily maintained after long-term use, and the limitation on the imaging assembly can be reduced.