866results about How to "Have diversity" patented technology

The invention discloses an image quality testing method based on a deep convolutional neural network. The image quality testing method based on the deep convolutional neural network comprises the following steps of firstly, establishing a sample set, establishing a deep convolutional neural network model, training the deep convolutional neural network model under different initial conditions, connecting the optimal deep convolutional neural network models obtained through the multiple times of training in parallel to obtain an image quality testing system, and using the obtained image quality testing system for testing images to be tested. According to the image quality testing method based on the deep convolutional neural network, characteristic learning is carried out by simulating the learning process of human brains, the problem that characteristics are difficult to select in an existing image quality testing method is solved, the contingency of predicting results is reduced, integration is high, generalization ability is strong, and testing effects are good.

The invention discloses a twin port calibration method for a nonlinearity vector network analyzer, relates to the technical field of microwave millimeter wave measurement and solves the problems that twin port calibration technology of existing nonlinearity vector network analyzers is single, and self-checking of a calibration result is difficult. Two ports of the nonlinearity vector network analyzer are utilized to perform open circuit, short circuit, loading, power, phase position and through calibrations respectively to obtain a nominal value and an actual measured value of a calibrated workpiece, eight error models are determined according to all obtained coefficients, and the calibration for the nonlinearity vector network analyzer is achieved. The twin port calibration method for the nonlinearity vector network analyzer is applicable to signal analysis and the most in-depth analysis for characteristics of non-linear devices.

The invention discloses a photoelectric control integrated input device switch module capable of generating twice touch click sounds, which comprises an upper cover, a central shaft active assembly, metal elastic sheets, a spring, optical elements, small hole diaphragms, a light emitting device, a light receiving device, a lower cover and a PCB. Two sides of a central key shaft are respectively provided with a sound production mechanism and a control mechanism, the sound production mechanism is located at one side of the central key shaft and comprises a convex part of an active sliding block and the metal elastic sheet for generating a key click sound; and the control mechanism comprises a convex part of the central key shaft, the optical element, the metal elastic sheet, a light baffling sheet whose edge is bent for a certain angle, and light emitting and receiving devices for controlling off or on of a light path. The module of the invention adopts an independent new click sound generation structure, on and off of the switch are not influenced, convenience, diversity and stability of design are achieved, and diverse different click sound can be created.

The invention discloses a deep belief network image recognition method based on Bayesian regularization and belongs to the field of artificial intelligence and machine learning. The deep belief network plays a more and more important role in the field of digital detection and image recognition. The invention provides a deep belief network based on Bayesian regularization on the basis of the network sparsity characteristic and changes of connection weights to solve the problem of overfitting in the training process of the deep belief network. By applying Bayesian regularization to the network training process, balance between error decreasing and weight increasing is effectively adjusted. The classification experiment of a digital script database proves effectiveness of the improved algorithm. An experimental result shows that in the deep belief network, the deep belief network image recognition method can effectively overcome the overfitting phenomenon and improve accuracy of digital recognition.

The invention relates to a three-dimensional printing method and device for tissue / organ chip integrated manufacturing. The three-dimensional printing method comprises the following steps: (1), designing the three-dimensional structural drawing of a three-dimensional chip, and converting the three-dimensional structural drawing into a slice layer graphic file format; (2), starting a three-dimensional printing device, and absorbing printing ink into all spraying heads; leading in slice layer graphic files; (3),respectively printing a main body material printing ink, a sacrificial material printing ink and different cell printing ink into pre-designed positions of a bottom plate system through the three-dimensional printing device; (4), repeating the step (3) and accomplish printing of a three-dimensional chip structure layer by layer and in an accumulating manner until the printing to a slice layer graphic file is finished; (5), heating or refrigerating the three-dimensional chip integrally printed, so as to enable a channel sacrificial material to be changed into a sol state; (6), using a pipette to absorb the channel sacrificial material in the sol state, removing the channel sacrificial material, and forming an integrated three-dimensional chip structure; (7), carrying out crosslinking on the cell printing ink material not molten, and carrying out culture medium perfusion.

The invention discloses a method for checking the reality of an ocean color remote sensing product. The method comprises the following steps: (1) reconstructing a long time sequence remote sensing data set; (2) computing and integrating based on the time-space variation characteristics of the remote sensing data set; (3) carrying out quality control and screening on a practical measurement data set for checking the remote sensing product; (4) checking generation of data sets by practical measurement-remote sensing data based on a time-space matching rule; (5) carrying out risk level division on the eligible matched data sets; and (6) computing indexes for checking the data reality of the remote sensing product. According to the method, the representativeness and the reasonability between the remote sensing product data and the checking practical measurement data can be checked, the matched data sets are screened and generated, and an evaluation method for different risk levels is obtained, so that the evaluation results are diversified and reasonable, the uncertain risk in the checking process is lowered, and the checking credibility is improved.

The invention discloses a construction method of an ecological sewage treatment and water restoration system. The system can be used for the miniature domestic sewage treatment, the aquatic environment rehabilitation of medium and small sized lakes and the landscaping water management and rehabilitation engineering, can achieve the target of building the self-regeneration and self-succession water ecological system and finally building the self-regeneration and self-succession water ecological system, and can build the landscape type sewage treatment plant. The method comprises the following steps: A. building a rainwater conflux and percolation water charging system; B. building an ET biochemical system; and C. gradually building an ecological balanced system of the aquatic animal food chain according to the relationship between the ecological chains of grass-fish, alga-fish, alga-debris-snail and debris-microorganism on the basis of finishing building the ET biochemical system.

The invention relates to a nano self-power-generation bionic flapping-wing air vehicle. The nano self-power-generation bionic flapping-wing air vehicle comprises nano self-power-generation bionic wings, a nano self-power-generation bionic tail wing, an intelligent controller, rechargeable batteries, sensors, a global positioning system, a signal receiving and transmitting device and a fuselage, wherein each of the nano self-power-generation bionic wings and the nano self-power-generation bionic tail wing comprises multi-joint driving rods; the multi-joint driving rods comprise wing arm front edge multi-joint driving rods, wing arm central pivot multi-joint driving rods and tail multi-joint driving rods; the multi-joint driving rods comprise joint drivers; each joint driver comprises a micro motor, a brake connection part and a pressure sensor module; the nano self-power-generation bionic wings, the nano self-power-generation bionic tail wing, the rechargeable batteries, the sensors, the global positioning system and the signal receiving and transmitting device are all connected with the intelligent controller. The nano self-power-generation bionic flapping-wing air vehicle has extremely important and wide application prospects in both civil and military fields.

An embodiment of the invention discloses a shooting method and electronic equipment, relates to the technical field of communication, and can solve the problem that an image shot by the electronic equipment is single. The method comprises the steps that under the condition that a first window and a second window are displayed, first input of a user to a target window is received, the first windowis used for indicating a first shooting preview picture of first electronic equipment, the second window is used for indicating a second shooting preview picture of second electronic equipment, and the target window comprises the first window and / or the second window; a target shooting preview image is updated in response to the first input, the target shooting preview image comprising a first shooting preview image and / or a second shooting preview image; a first camera of the first electronic equipment and a second camera of the second electronic equipment are controlled to execute shooting operation respectively based on the updated target shooting preview picture, and a target image is output, wherein the target image is an image obtained by synthesizing an image corresponding to the first shooting preview image and an image corresponding to the second shooting preview image.

The embodiment of the invention discloses an image restoration method and device and a storage medium. According to the embodiment of the invention, the to-be-restored image is obtained; A to-be-repaired area and a non-repaired area can be determined from the to-be-repaired image; performing feature extraction on the non-repaired area based on different receptive domains and spatial resolutions, then generating textures of the to-be-repaired area according to the extracted feature information of multiple scales, and filling the generated textures into the to-be-repaired area in the to-be-repaired image to obtain a repaired image; According to the scheme, the target with the high structural property or the large area can be well repaired, the authenticity of the repaired area is greatly enhanced, and the repairing effect is improved.

The invention discloses an underwater acoustic anti-Doppler multicarrier modulation and demodulation method based on linear frequency modulation. The method comprises the following steps: step one, transmitting terminal multicarrier modulation is performed; step two, Doppler frequency shift factor estimation is performed; and step three, receiving terminal demodulation is performed. The invention also discloses a modulation and demodulation device of the underwater acoustic anti-Doppler multicarrier modulation and demodulation method based on linear frequency modulation. The device comprises a transmitting device and a receiving device which are interconnected. The method has advantages of being high in anti-Doppler capability, wide in applicability, diverse in realization method, low in cost, easy to realize, high in efficiency, etc.

The disclosure relates to a method and an apparatus for changing emoticons at a chat interface, and terminal equipment. The method includes: whether an instant message received by a chat interface of an instant messenger includes a first emoticon which allows to be changed is detected; if the instant message includes the first emoticon, a first emotion-score list is checked, and a first score corresponding to the first emoticon is determined; a second emotion-score list is checked, a second score corresponding to a second emoticon is determined, and the second emoticon is an emoticon used for changing the first emoticon; the first score and the second score are accumulated, and a score result is obtained; and the first-emotion-score list is checked, and the first emoticon is changed to a form corresponding to the score result. According to the technical scheme, the emoticons in the instant chat can be provided with the score result capable of expressing the emotion, the received emoticons are dynamically changed according to the score result, and users can perceive the dynamic change of the emotions.

The invention discloses a distributed flexible pressure sensor based on an electrically active polymer. The distributed flexible pressure sensor comprises a sensor array, a switching circuit, an interface circuit and a computer. The sensor array is used for preparing a corresponding array according to the size of a measuring plane and inputting pressure. The switching circuit is used for converting a sensed capacitance value into a voltage value. The interface circuit is used for conducting line-rank scanning, filtering, amplifying and A / D conversion on an input signal and converting an analog voltage signal into a digital voltage signal capable of being read by the computer. The computer is used for collecting, analyzing, processing and displaying the data input by the interface circuit. The flexible pressure sensor can be widely applied to the fields of human health monitoring, physical training, medical rehabilitation, biomechanics and safety monitoring and the like.

The invention discloses a vertical beam falling-off prevention system. The system comprises three parts: the first part are square mesh energy consumption falling-off prevention beams placed and connected between a beam bottom and a capping beam, the second part are sliding cylinder connecting rod devices connected with two side surfaces of a bridge end above a pier and abutment, and a third part are anti-collision pieces placed at connection parts of bridge web boards. The invention relates to a vertical limit energy consumption beam falling-off prevention protective device for a bridge under earthquake. The protective device is simple in manufacturing, convenient to assemble and light in self weight. During normal operation, the device can be freely stretched out and drawn back along with the structure, under the action of the earthquake, the square mesh energy consumption falling-off prevention beams and sliding connecting rods drive shape memory alloy wires to be stretched for bearing force and wasting energy so as to effectively absorb earthquake destroy energy, and can limit the bridge to generate large displacement, so that under the action of strong shock, the damage due to the falling off of beams is effectively prevented, and the anti-collision pieces at the beam end can prevent the beam from collision. The system provided by the invention can effectively solve the problem that vertical displacement safety, continuous energy consumption capacity and falling off beam capacities of railway bridges under earthquake can not be ensured.

Present invention discloses method for dynamical generating caption image data and data stream. It contains 1, visiting caption user describing record obtaining source data, 2, proceed caption attributive resolving to source data of caption user record, obtaining generating caption data rule or listing promise, 3, utilizing step 2 obtained rule or listing promise automatically drawing generating caption image using computer graphic equipment through calculation, and obtaining caption each frame picture image data, single frame image continuously constituting a caption image data sequence, 4, encoding caption image data in step 3 and packaging to data stream fitting digit caption standard data stream. Present invention can enrich digit caption display form, raise data generation and broadcast mobility, diversity and self control performance without manual operation.

The invention discloses a construction method of a wifi indoor positioning fingerprint database based on a generative adversarial network. The method includes the steps of 1, obtaining the channel state information, namely CSI data, for measuring the channel condition through measurement; 2, respectively processing CSI data of each position into corresponding amplitude phase characteristic graph to preliminarily construct a fingerprint database; 3, sending the obtained amplitude phase characteristic graphs to the generative adversarial network for training to generate more amplitude phase characteristic graphs, so as to construct a wifi indoor positioning fingerprint database with sufficient samples. The signal sampling quantity is reduced by using the generative adversarial network to generate amplitude phase characteristic graphs with different positions, so that the problem that a library cannot be effectively built when the CSI data is insufficient is solved, the measurement efficiency is improved, and the method is simple and convenient to implement.

The invention relates to an active learning sample selection strategy integrated with a confidence criterion and a diversity criterion. The active learning sample selection strategy comprises the following steps: training a model Mt based on an existing labeled data set DL; predicting a current unlabelled data set DU by using the Mt to obtain a predicted vector set Pt; calculating an information entropy of each sample according to the Pt, and selecting front K samples each having a largest entropy; extracting feature representations of K unlabelled samples according to the Mt to obtain a feature vector set Ft; performing density peaks clustering on the Ft, respectively selecting corresponding proportion and number of samples from a center of a cluster generated by the density peaks clustering, and an edge point and an outlier of the cluster, handing the samples to an expert for labeling, adding the labeled data set DL, and simultaneously deleting corresponding samples from the unlabelled data set DU; updating the Mt by using the current labeled data set DL to obtain Mt + 1; and repeating the above steps till labeling of all samples is ended or reaches to a designated number of iteration times to complete a whole algorithm flow.

The invention discloses a code morphing-based binary code obfuscation method. The method comprises the following steps: 1, determining a morphing template library; 2, detecting a PE (Portable Execute) file; 3, locating a key code segment; 4, disassembling the key code segment to obtain an assembly instruction sequence, identifying a circular structure in the code segment, and calculating the weight of each instruction in the key code segment; 5, morphing all assembly instructions in the assembly instruction sequence; 6, slicing the assembly instruction sequence and re-sequencing; 7, assembling a new assembly instruction sequence generated in the step 6, and converting into a binary code; 8, reconstructing the PE file. In the method, a computer system is used for obfuscating the executable binary code file under a windows system, so that high protection strength is achieved, and extension becomes easy.

The invention provides a method for patterning a metal surface. The method comprises the following steps of: firstly, cleaning a metal substrate by chemical or physical treatment and polishing the metal substrate; secondly, performing anodic oxidation treatment on the metal substrate so as to form a porous oxide film on the surface of the metal substrate, wherein the porous oxide film is provided with a plurality of pores with the pore diameters of between 20 and 400 nanometers; thirdly, providing a pattern layer and absorbing ink particles into the pores of the porous oxide film in a printing mode, wherein the particle sizes of the ink particles are between 10 and 300 nanometers; and lastly, sealing the pores so as to form a pore sealing layer which is continuously covered on the surface of the porous oxide film.

The invention discloses a dish basket assembly for a dish washing machine and the dish washing machine with the same. The dish basket assembly comprises a frame and movable shelves. The frame comprises a frame bottom wall and frame side walls. The movable shelves are arranged in the frame. Each movable shelf comprises an installation rod and multiple stop rods which are fixed to the installation rod, wherein the multiple stop rods are arranged in the length direction of the installation rod in a spaced mode, and the installation rod is connected with the frame in a pivoting mode. According to the dish basket assembly for the dish washing machine, due to the fact that the movable shelves are arranged in the frame in a pivoting mode, the placing angles of dishes can be adjusted by changing the arrangement angles of the movable shelves, in this way, tableware can be at the optimum cleaning angle, and the tableware can be washed to be cleaner; in addition, due to the fact that the arrangement angles of the movable shelves can be adjusted, tableware in various specifications can be placed in the frame, and flexibility and applicability of the dish basket assembly are achieved.

The invention discloses a process for treating wastewater generated in a solar silicon wafer production process and relates to wastewater treatment in the field of the electronic semiconductor industry. The process comprises the following steps of: pretreating wastewater in a preliminary sedimentation tank; carrying out Fenton and iron-carbon micro-electrolysis combined oxidization reaction on the wastewater in an iron-carbon micro-electrolysis reactor; after the reaction, feeding the wastewater into a coagulation-sedimentation integrated pool, feeding PAC (Poly Aluminium Chloride) and PAM (Poly Acryl Amide) to arouse the coagu-flocculation reaction of the wastewater, and after the reaction, performing sludge-water separation on the wastewater; causing the supernate to flow to a hydrolysis acidification pool, and then causing the wastewater to flow to a contact oxidation tank; causing the mixed liquid to flow to a secondary sedimentation tank for sludge-water separation, while causing the supernate to flow to a deflection sedimentation pool, and feeding PAC for purifying water so that the supernate is up to the stand and can be discharged after sedimentation; finally, performing concentration treatment on sludge. After non-degradable organic wastewater is treated by the process provided by the invention, the BOD5 (Biological Oxygen Demand) / CODcr (Chemical Oxygen Demand) ratio of the wastewater is obviously increased.

The invention discloses a display area control method and system and a mobile terminal. The display area control method comprises the following steps: receiving the adjustment request of a display area; acquiring the target shape of the display area which needs to be adjusted; controlling the content needing to be displayed in the display area with the target shaped to display. The display area control system comprises a request receiving module, an acquiring module and a control module; the request receiving module is used for receiving the adjustment request of the display area; the acquiring module is used for acquiring the target shape of the display area which needs to be adjusted; and the control module is used for controlling the content needing to be displayed in the display area with the target shaped to display. The mobile terminal comprises the display area control system. Due to the adoption of the technical scheme, the problems in the prior art that the display areas of various electronic equipment are fixed and the shapes of the display areas can not be adjusted are solved.

The invention discloses a preparation method of a porous 5A molecular sieve for straight-chain alkane adsorption and belongs to the technical field of molecular sieve adsorbents. The preparation method comprises the following steps of adding ethyl orthosilicate into a sodium hydroxide aqueous solution, carrying out stirring for some time, fully dissolving sodium metaaluminate in water, fast mixing the two solutions, carrying out violent stirring to obtain milky colloid, carrying out crystallization on the colloid at a temperature of 80-95 DEG C for 4-6h, carrying out cooling, carrying out filtration, carrying out washing until a pH value is less than 9, carrying out drying by an oven to obtain 4A molecular sieve raw powder, carrying out ion exchange on the 4A molecular sieve raw powder and a calcium chloride solution to obtain the 5A molecular sieve. The 5A molecular sieve has n-decane static saturated adsorption capacity of 0.538g / g and n-pentdecne static saturated adsorption capacity of 0.647g / g. The 5A molecular sieve is prepared from an organic silicon source ethyl orthosilicate replacing an inorganic silicon source sodium silicate so that the channels of the 5A molecular sieve have diversity. The 5A molecular sieve has a good saturation adsorption amount of straight-chain alkanes.

The invention discloses a fraud group identification system. The system comprises a data source, a data loading module, a data storage module, a comprehensive analysis module and an application module, wherein the data source is used for providing original log data; the data loading module is connected with the data source and is used for acquiring and preprocessing the original log data; the datastorage module is used for storing the original log data and preprocessed intermediate result data; the comprehensive analysis module is internally provided with a fraud group identification module;the fraud group identification module is connected with the data storage module; the fraud group identification module is used for identifying a fraud group and storing fraud group information in thedata storage module; and the application module is connected with the data storage module and is used for displaying the fraud group information. According to the fraud group identification system, through quantitative analysis for online data, the fraud group is discovered, the fraud group can be comprehensively identified and the possible fraud group in future can be identified.

The invention belongs to the technical field of intelligent manufacturing equipment, in particular to intelligent manufacturing unit system integration design teaching equipment, which comprises multiple function units arranged independently for implementing storing, processing, detecting or sorting processes of workpieces. The multiple function units arranged independently comprise an execution unit, a tool unit, a storage unit, a processing unit, a polishing unit, a detecting unit, a sorting unit and a main control unit which can be combined in overall arrangement through connection according to the needs; the main control unit provides a power source and an air source for other function units, and connection is realized by utilizing quick-connecting cables; the execution unit, the storage unit, the processing unit, the polishing unit, the detecting unit and the sorting unit are controlled by a controller of the main control unit to act through industrial ethernet. During a use process, the equipment can show the advantages of an intelligent control technology, is increased in use flexibility, has the advantage of diversity in a teaching process, and can meet the teaching requirements on individuality and specificity.

The invention relates to a multi-ejector parallel type solar ejecting refrigeration unit which utilizes solar energy to supply cold air, which comprises a solar hot-water loop, an ejecting refrigerant loop and an air conditioner water loop. The unit solves the problems that the prior ejecting refrigeration system is only provided with an ejector, the structural dimension of the ejector is fixed and the efficiency of an ejector system is drastically declined directly because operation condition is derivated from design condition. The unit is arranged between a generator and a condenser, and is provided with two or more than two ejectors with different structural dimensions in parallel. The multiple ejectors can meet high-efficiency operation of variable working conditions; and the unit can automatically or manually switch to a proper ejector for working according to solar radiation or weather changing conditions, further ensure that the system operates in refrigeration season with high efficiency, can improve the operation efficiency of the solar ejector system, and realize the aim of supplying the cold air with high efficiency and reasonability, thereby improving the utilization rate of solar energy, saving energy, relieving the pressure of electricity grid at the peak time, and improving the light-heat utilization efficiency.

The invention provides a Chinese reading ability evaluation method, which comprises the following steps of: obtaining user information of a user, and calling an initial test question group corresponding to the user information in an evaluation question bank for the user to test; Calculating an initial capability value of the user according to a test result, obtaining an evaluation test paper corresponding to the initial capability value, and sending the evaluation test paper to a terminal where the user is located for the user to perform evaluation, the evaluation test paper comprising test question groups of multiple question types of continuous texts and discontinuous texts; And based on the project response theory, calculating a reading capability value of the user for answering the evaluation test paper by utilizing a preset high-order question group model. The invention also provides a Chinese reading ability evaluation device and a readable storage medium. According to the methodand the system, the user is tested according to the test papers of various question types of the continuous text and the discontinuous text, the item response of the user for answering the test papers is evaluated through the high-order question group model, and the question types have diversity and can more accurately and comprehensively inspect the Chinese reading capability of the user.

Disclosed is a book recommendation method and system based on social networking. The method includes the steps of 1, extracting a user's interactive messages with other users in a social network, establishing a plurality of interactive type friend groups for the user, and dividing the other users, who successfully interact with the user, into the different interactive type friend groups according to the interactive types; 2, calculating the number of successful interactions between the user and each friend in the interactive type friend groups, selecting several front friends, who successfully interact with the user for the most times, from each interactive type friend group, and recommending books, read by the several friends for the most times, to the user. The method and system belonging to the technical field of network communications has the advantages that books can be individually recommended according to interactions of users in the social network.

The invention relates to a similarity propagation and popularity dimensionality reduction based mixed recommendation method. According to the similarity propagation and popularity dimensionality reduction based mixed recommendation method, sparse data are processed in two phases; firstly, neighbors of the sparse data are expanded due to constant iteration of similar matrixes of users, resources and Tags through a similarity propagation method and accordingly elements for zero are filled; then a score algorithm in a search engine is introduced to calculate the Tag popularity in consideration of the problem that original data is provided with meaningless rubbish Tags, the tags with the popularity smaller than a certain threshold value are deleted to simplify data to perform dimensionality reduction on the matrix; recommendation results are diversified and the sparsity and cold starting problem can be relieved to some extent due to the fact that the recommendation based on contents and the collaborative filtering recommendation are combined. The similarity propagation and popularity dimensionality reduction based mixed recommendation method has the advantages of solving the problem of data sparsity in the individual recommendation process and being high in recommendation result accuracy, high in accuracy and high in reliability.

The invention relates to the field of synthetic production of cosmetics and aims to provide strippable water-based non-irritation nail polish. The strippable water-based non-irritation nail polish comprises deionized water, water-soluble modified polyurethane, chitosan, a wetting agent, a dispersing agent, an antifoaming agent, a rheological agent, water-soluble environment-friendly color paste and other additives. Water is used as a solvent, and no or only a small amount of absolute ethyl alcohol is added to be used as an additive, so that the nail polish has no paint smell of traditional oil-soluble nail polish or alcohol smell of alcohol-soluble nail polish and contains no alkali-associative thickener, and the whole nail polish system is neutral and causes no damage to the skin, nails and clothes.