103results about How to "Improvement effect" patented technology

The invention discloses a terminal voice noise reducing method and a terminal. The terminal comprises a first microphone and a second microphone. The terminal voice noise reducing method comprises the following steps of: judging whether the ambient environment is quiet or noisy by a sound pressure level or signal to noise ratio of the noise at the position in which the terminal is located, if a sound pressure level is lower than the preset sound pressure level threshold value or the signal to noise ratio is higher than a preset signal-to-noise ratio threshold value according to a detection result, judging the ambient environment of the terminal to be the quiet environment, and using the first microphone to carry out noise reduction treatment on voice. A condition that a sound sent by a user does not cause loss because of the noise reduction treatment is guaranteed, and a talking voice effect is improved.

The invention, which belongs to the field of image processing, discloses an image enhancement processing method and device. The method comprises: a specified image is processed by using a first regression network model to obtain a content feature map of the specified image; a gradient map of the specified image is processed by using a second regression network model to obtain an edge feature map of the specified image; and with a fusion processing model, the content feature map and the edge feature map are fused to obtain a restored image. With the image enhancement processing method, the processing effect is improved; and the processing efficiency is enhanced.

The invention provides a method and a device for extracting a text from an image, and the method for extracting the text from the image comprises the following steps of A, conducting binaryzation processing for the image to obtain all connected domains of the image; B, filtering the connected domains which cannot satisfy a first statistics characteristics, wherein the first statistics characteristics are the statistics characteristics which are obtained after statically learning description characteristics of the extracted connected domain in a well-labeled sample and belong to the text connected domain; and C, extracting the text from the image from each connected domain after being filtered. Through the method, the accuracy for extracting the text from the image can be greatly improved.

The invention relates to a mixed unit of a foam metal ball pipe assembly. The mixed unit consists of an exhaust pipe, a nozzle, a connecting flange, a welding spot, a fastening outer sleeve, a fastening inner sleeve, a metal fiber braided pipe and a foam metal ball. The mixed unit is characterized in that: the fastening outer sleeve and the fastening inner sleeve are used for fixing the metal fiber braided pipe and welded into a whole by using the connecting flange; the inside diameter of the fastening inner sleeve is D1, the diameter of the foam metal ball is D2, and the ratio D2 / D1 of the diameter D2 of the ball to the sleeve diameter D1 is 1.5 to 2.5; the ratio L2 / L1 of the length L2 of the fastening outer sleeve to the length L1 of the fastening inner sleeve is 3 to 5 millimeters; L2 is 300 millimeters; and the mixed unit is arranged at the downstream of the nozzle for a urea reducer. The foam metal ball is wrapped by adopting the metal fiber pipe to form the mixed unit of the concentric cylindrical foam metal ball pipe assembly for the urea reducer, a catalyst is coated on the foam metal ball, and the mixed unit is applied to mixed decomposition of the urea reducer; and the mixed unit has simple structure, is efficient and durable, and is applied in an over national 4 post treatment device.

An image processing method based on a neural network, an image processing apparatus, and a computer-readable storage medium are provided. The image processing method comprises the following steps of:obtaining an input image; classifying the input image by using neural network to obtain multiple category labels, wherein the neural network comprises a first sub-neural network and a second sub-neural network, The second sub-neural network comprises a plurality of branch networks, at least two branch networks of the plurality of branch networks are arranged in parallel, and the step of classifying the input image by using neural network to obtain multiple category labels comprises the following steps: extracting basic feature information of the input images by the first sub-neural network toobtain a plurality of first feature images; Dividing a plurality of first feature maps to obtain a plurality of first feature map groups corresponding to the plurality of branch networks one by one;respectively processing a plurality of first feature map groups by a plurality of branch networks to obtain a plurality of class tags.

The invention discloses a remote sensing image semantic segmentation method based on a pyramid pooling multistage feature fusion network. The method comprises steps of employing ResNet to extract features, respectively introducing the features extracted at each stage of ResNet into a spatial pyramid pooling structure to extract the multi-scale information of a target, and introducing a double attention module to the last part of feature extraction, and a multi-level feature fusion strategy is adopted to perform feature fusion on the pooled features and the features acquired by the double attention module so that refined classification of the remote sensing images can be realized.

The invention discloses a dark channel sky area defogging method based on threshold segmentation optimization, and mainly solves the situation of poor processing effect of the sky area of an image of a foggy day in the prior art. The realization steps of the dark channel sky area defogging method based on threshold segmentation optimization are that (1) a randomly selected foggy image including the sky area is inputted, and the dark channel value and the atmospheric light value of the image are calculated in turn; (2) the transmission graph of the foggy image is solved; (3) orienting filtering is performed on the transmission graph so that the optimized transmission graph is obtained; (4) the sky area is marked; and (5) transmittance of the sky area is amplified so that the defogged image is restored. Fog information of non-sky areas of the image of the foggy day can be effectively eliminated, and the sky area processing effect can be optimized so that the phenomenon of color distortion and halation of the sky area can be avoided, and thus the dark channel sky area defogging method based on threshold segmentation optimization can be used for image enhancement and video monitoring.

The invention provides a data cleaning method and device. According to the method, multiple cleaning rules are set according to different data features in advance; and when target data needs to be cleaned, cleaning rule matching is performed according to the data features of the target data, and then the matching cleaning rules are utilized to clean the target data. Therefore, it is guaranteed that the cleaning rules adapt to the data features, the target data can be cleaned in a more targeted mode, more dirty data can be effectively cleaned, meanwhile, the probability of recognizing clean data as dirty data by mistake is lowered, and the cleaning effect is improved.

The invention discloses a crowd exceptional event detection method. The method comprises the steps that a first image and a second image are acquired; multiple optical flow points of the first image are calculated according to the first image and the second image; an optical flow point adjacency matrix is calculated according to a horizontal coordinate relation and a vertical coordinate relation among all the optical flow points; the optical flow points are divided into corresponding region sets according to the optical flow point adjacency matrix and clustering conditions; each region set isdivided into at least one region subset according to a distance relation among the optical flow points; all the region subsets are corrected according to historical grouping information, and correctedsubsets are obtained; and whether a crowd exceptional event occurs is judged according to an average motion distance of the optical flow points in each corrected subset. According to the method, optical flow is adopted to represent motion features of a scene, cluster grouping is performed on the optical flow points, then optical flow information in each group is calculated to perform statisticalanalysis on motion conditions of the corresponding group, therefore, an exceptional event occurring in a crowd can be detected, and the method can be suitable for different scenes and has good instantaneity.

The invention discloses a miniature combine harvester which is good in harvesting quality. The miniature combine harvester comprises a body and a cleaning mechanism arranged on the body. The miniature combine harvester is characterized in that the cleaning mechanism comprises a cyclone separator arranged on the combine harvester, a shell of the cyclone separator is divided into a cylindrical part and a conical part from top to bottom in the axial direction, and helical impellers are axially arranged inside the cyclone separator; a vertically-upward air outlet and a transverse feed port are formed in the cylindrical part, a discharge port is formed in the lower end of the conical part, and vent holes are formed in the shell wall, close to the discharge port, of the conical part; the feed port is connected with an output port of a threshing device, and the air outlet is connected with an input port of the threshing device. Some cereal grain which is not plump and moves along with airflow can be secondarily cleaned, and the cereal grain is prevented from being discharged as impurities. Accuracy of impurity removal is improved, and the cleaning effect is improved.

The invention provides a preparation method of recycled aggregate for inhibiting an alkali-aggregate reaction. The preparation method is prepared from the following steps (1) the recycled aggregate iscollected, specifically, recycled concrete is crushed and sieved, and thus the recycled aggregate with the concrete required size is prepared; (2) silica fume slurry is prepared, specifically, according to the water-solid mass ratio of 5:1 to 15:1, silica fume is dissolved in water, and the silica fume slurry is prepared; (3) the silica fume slurry is subjected to vacuum suction to prepare the recycled aggregate, specifically, the collected and prepared recycled aggregate and the silica fume slurry are jointly placed into a vacuum container with the negative pressure for 6-18 h, and it is ensured that the silica fume slurry is fully sucked into recycled aggregate attached slurry body; and (4) the recycled aggregate with the sucked silica fume is dried, specifically, the recycled aggregatesubjected to vacuum sucking of the silica fume slurry is dried, and thus the recycled aggregate for effectively inhibiting alkali activity is obtained. According to the preparation method, a technology is simple, the cost is low, meanwhile, the recycled aggregate with the alkali activity can be efficiently utilized, and the environmental pressure is relieved.

The invention relates to a new additive for iron ore sinter, which can greatly improve sintering granulating effect, increase sintering mineralization rate and strength, lower production cost, and improve production efficiency. The invention is characterized in that the new additive for iron ore sinter is composed of 60% of calcium oxide, 2% of silicon dioxide, 2% of Al2O3, 4% of magnesia, 1% of ferric oxide, 1% of calcium chloride and 30% of calcium carbonate. The above materials are uniformly mixed, and then free moisture is dried by a vapor dryer, the water content is 49%, finally, the materials fine grinded by a globe mill to form the new additive for iron ore sinter. The invention can improve the production of sintering granulating 3mm to 5mm granulation pellet from 24% to 36.7%, the fine granulation is reduced obviously, the sintering granulating effect is improved significantly, the sintering mineralization rate and strength are increased notably, thereby lowering production cost and improving production efficiency.

The present invention provides a link load balancing method and device The method comprises: obtaining the domain name carried in a DNS response message when a load balancing device receives the DNS response message sent by a DNS server, and the DNS response message is used for replying the message of an IP address corresponding to the domain name required by the DNS request message; determining whether the domain name matches a configured specific domain name in advance or not; if domain name matches a configured specific domain name in advance, generating a dynamic item, wherein the dynamic item includes the IP address corresponding to the domain name carried by the DNS response message; when receiving the data message sent by a user terminal, inquiring the dynamic item according to the object IP address of the data message; and if there is provided an IP address being the same as the object IP address of the data message in the dynamic item, forwarding the data message through a specific link. Through adoption of the embodiment of the invention, the link load balancing method and device are able to optimize the load balancing effect.

The invention relates to equipment and method for preparing a large-area high-density nuclear track nano-pore membrane. The equipment comprises an etching generation device, wherein the etching generation device comprises a clamping part and a liquid holding part; the clamping part is used for clamping a latent track film and is fixed on the liquid holding part filled with etching liquid; an intermittent ultrasonic action is provided by utilizing an ultrasonic generator, and an etching product remained in track holes can be peeled in time; the latent track film is subjected to radiating sensitization by utilizing an ultraviolet lamp, and the etching speed is accelerated. According to the equipment and method, uniform etching of the film can be realized; the etching product is peeled by utilizing ultrasonic waves, so that the etching liquid and the latent track can be sufficiently contacted, and the etching speed of the track can be accelerated; the film is protected against the broken risk due to long-time vibration of the ultraviolet waves by adopting the intermittent ultrasonic action; and the track nano-porous film has excellent reflection resistance, and has wide application range due to the large area and high pore density.

The invention provides a method for remediating soil contaminated with volatile organic compounds. The method comprises the steps that the contaminated soil is subjected to crushing and screening treatment, activated aluminum oxide is added into soil particles according to the proportion of 10g: 1kg, sodium dodecyl benzene sulfonate is added into the soil particles according to the proportion of 20g:1kg, and 0.5-1 L of purified water is added into the soil particles to obtain a soil mixture; and then the soil mixture is placed into a desorber, provided with a heating unit and a compressing unit, of a vacuum desorption device, compressing and heating desorption is conducted, the volatile organic compounds in the soil are desorbed under heating and squeezing effects, and purified soil is obtained. The method is easy in implementing process, easy to operate, free of secondary pollution in the whole process, high in desorption efficiency, and suitable for treating heavily organic contaminated soil.

Integrated plant enzyme relates to the field of enzymes and comprises, by weight, 35-45 parts of pancreatin replacement enzyme, 0.5-1.5 parts of protease 3.0, 0.5-1.5 parts of protease 4.5, 0.5-1.5 parts of protease 6.0, 0.5-1.5 parts of bromelain, 8-10 parts of carbohydrate hydrolytic enzyme and 0.5-1.5 parts of fat hydrolytic enzyme, wherein the protease 3.0 is protease with the optimum pH value being 3; the protease 4.5 is protease with the optimum pH value being 4.5; and the protease 6.0 is protease with the optimum pH value being 6. The integrated plant enzyme can take effect on human bodies with different pH values, has a full effect of improving the function of human body, has the function of pancreatin, and has an integrated effect of improving the function of human body. In addition, the invention also designs the application of the integrated plant enzyme in food and health food.

The invention discloses an electromagnetic heating device and a heating control method thereof. The heating control method of the electromagnetic heating device comprises the following steps of sampling a current of a switch tube in the electromagnetic heating device to generate a voltage sampling signal; acquiring a current heating power of the electromagnetic heating device, and judging a heating state of the electromagnetic heating device according to the current heating power; selecting corresponding amplification rate to amplify the voltage sampling signal according to the heating state of the electromagnetic heating device, and acquiring a current sampling value according to the amplified voltage sampling signal; and adjusting the heating power of the electromagnetic heating device by comparing the current sampling value with a target current value. According to the method, different amplification rates can be selected to sample the current according to the heating state of the electromagnetic heating device, so that the purpose of preventing power fluctuation of the electromagnetic heating device is achieved, the cooking effect is improved, and serious heat generation of the switch tube can be prevented.

An efficient torretial flow prompted phase change method for separating the oil from water includes such steps as filtering the polluted industrial oil, heating to 40-70 deg.C, loading in torretial flow generator where water changes from liquid phase to vapour phase, and oil-water separating in the oil-water separator. Its advantages are high effect and speed, and low energy consumption and cost.

The invention provides a tongue image segmentation method based on the threshold technology and gray scale projection, and the method comprises the following steps: S1, transforming a hue component ofan obtained tongue image in an HSI color space, so that there is a difference in hue between the hue component of a tongue and an upper lip and its neighbor tissues after transformation; S2, employing the image threshold segmenting technology on the transformed hue component to obtain a binary segmentation result, and performing a morphological operation of the result to extract an initial tongueregion including the true tongue and the upper lip; S3, using the gray projection technique to find an image line corresponding to the boundary of a tongue root, removing the false tongue regions through the image line, such as the false region of the upper lip and the gap between the tongue root and the upper lip, correcting the tongue region, and taking the boundary of the tongue region as an initial tongue contour; S4, carrying out the smoothing processing of the tongue contour through using an active contour model to obtain a final segmentation result of the tongue image. The method remarkably improves the tongue image segmentation effect.

A method includes following steps. The difference between the original pixel measured in spatial neighborhood of the pixel to be interpolated and the corresponding motion compensation pixel. The saiddifference value is utilized to estimate the compensation error. The estimated error is used to correct the motion compensation. The devices for executing the method includes the error detector utilized to detect and measure the differences between several original pixels and the corresponding motion compensation pixels, and the correcting compensator utilized to estimate the said differences andcorrect the motion compensation. The invention possesses the advantages of less operation quantity and good compensation effect.

The invention provides a head-mounted device of an eyeball tracking device having multi-layer free-form optical waveguide. The head-mounted display device comprises a main body casing, an eyeball tracking device having a multi-layer free-form optical waveguide, a main board and a power source. The eyeball tracking device having the multi-layer free-form optical waveguide includes a lens, an infrared light source and an eyeball imaging sensor. The main board includes an eyeball tracking calculation module and a device control module. Through the technical solution of the invention, the problemof low integration of the device with the eyeball tracking device is solved, a new scheme is provided for the simplified design and the appearance expansion, the eyeball tracking and positioning speedof the device is fast, and the tracking effect is good.

The invention discloses a metal mine downhole microseismic monitoring system which comprises at least one downhole network multi-channel collector arranged in a roadway of a downhole monitored section. Every downhole network multi-channel collector comprises a plurality of difference input channels for analog signal collection, every difference input channel is respectively connected with a microseismic signal collecting device, every microseismic signal collecting device is fixedly distributed on downhole roadway side walls in different directions, and output ends of the downhole network multi-channel collectors are connected with a ground data receiving processing platform sequentially through a transmission local area network and a concentrator local area network. The metal mine downhole microseismic monitoring system has the advantages of performing microseismic real-time on-line monitoring and analysis, preventing downhole accidents, optimizing blasting and ore falling effect in the recovery process and mastering vibration conditions of surrounding device systems and the like and is suitable for being widely applied to the technical field of microseismic monitoring.

The invention belongs to the technical field of liquid component mixing and particularly relates to a multi-component atomizing and mixing technology. According to the technical scheme, a multi-component atomizing and mixing device comprises a plurality of atomizing devices atomizing liquid components, and a mixing device for mixing fog drops after the various liquid components are atomized. A multi-component atomizing and mixing method includes the steps that A, the various liquid components are atomized and dispersed into the fine and small fog drops; and B, the fine and small fog drops formed by atomizing and dispersing the various components are mixed in one space. According to the multi-component atomizing and mixing technology, the structure is simple, and atomizing and mixing of the liquid components under the controllable condition of factor variables can be achieved.

The invention relates to precision casting wax. The precision casting wax comprises the following ingredients, by weight, 80-90% of PE wax, 4-10% of polystyrene, 4-10% of eleaostearic acid and 0.2-0.4% of white carbon black, and the mass percent sum of the precision casting wax is 100wt%. During the CAD precise processing of the precision casting wax, re-fusion and wire drawing phenomena in a small-sized scope are reduced obviously, and the effects of the precision casting wax with PE wax as base wax and used for CAD precise processing are improved obviously. The internal components of the precision casting wax are uniform, the shrinkage of each part is uniform, and the wax block quality is raised.

The invention relates to a fullerenes complex and an application of the fullerenes complex in removing free radical in smoke. A metallic organic matrix compound crystalline material is taken as a carrier, fullerenes molecules are absorbed in the surface or pores, and the obtained complex is used as the material for removing the free radical in the smoke. The result shows that the free radical in the smoke is well removed; the free radical in the cigarette smoke can be removed, and the free radical in vehicle exhaust, industrial waste gas and indoor gas can be removed as well.

The invention relates to a far-infrared sauna bodybuilding system for non-therapeutic purposes based on physical sign parameters and a use method thereof, belonging to the technical field of electronics and solving the problems that the traditional sauna body-building system cannot accurately supply temperature and humidity environments moderate for a human body, is relatively troublesome to use and the like. The far-infrared sauna body-building system based on physical sign parameters comprises a far-infrared sauna room, a health management unit, an identity authentication unit connected with the health management unit, a physical sign parameter testing unit, a display unit and a temperature and humidity control unit. The use method of the far-infrared sauna body-building system based on physical sign parameters comprises the steps of: (1) starting and initializing a system; (2) inputting and identifying identity information; (3) testing physical parameters; (4) analyzing and generating a body-building instruction scheme; (5) executing the instruction scheme; (6) carrying out sauna showering; and (7) closing the system. The sauna bodybuilding system has the advantages of convenience for use and capability of providing a moderate temperature / humidity-controlled environment.

A medical text OCR method and system are disclosed in the field of optical character recognition and medicine. The medical text OCR method comprises the following steps: preliminarily recognizing a medical text picture as a text file; Classify the text file according to the predefined text type, and determine the text type of the text file; The text file is accurately recognized by a special OCR recognizer corresponding to the text type to which the text file belongs. A medical text OCR scheme is proposed, in which picture contents are recognized by using a universal OCR recognizer, thereby distributing the preliminary recognized picture contents to different dedicated OCR recognizers for accurate recognition, thereby improving the effect of medical document OCR recognition.

The invention relates to a stomach nourishing composition for regulating the spleen and the stomach. The stomach nourishing composition is prepared from, by weight, 20-43 parts of herba rhodiolae extract, 50-68 parts of herba dendrobii extract, 20-35 parts of fresh ginger extract and 10-15 parts of honey. The stomach nourishing composition can strengthen the spleen, relieve dyspepsia and harmonize the stomach and has the obvious improvement effect on spleen-stomach deficiency type gastritis, and in the composition, the herba rhodiolae extract, the herba dendrobii extract and the fresh ginger extract have the synergistic effect on treatment of the spleen-stomach deficiency type gastritis.

The invention discloses an orthopedics-department joint replacement operation system. A control table comprises a display, and an operation plan is stored in the control table; the operation plan comprises a to-be-operated joint three-dimensional graph, a prosthesis model and three-dimensional graph and the position corresponding relationship and the whole force line relationship between a prosthesis and the to-be-operated joint; a to-be-operated-joint real-skeleton-local geometric curved surface obtained according to a scanner is matched with the to-be-operated joint three-dimensional graph in the operation plan through the control table, the actual position of the joint three-dimensional graph in the operation plan corresponding to a to-be-operated-joint real skeleton is traced, determined and traced in real time relying on infrared optical navigation, and the specific position of the prosthesis corresponding to a skeleton local is displayed on the to-be-operated joint three-dimensional graph in the operation plan. According to the orthopedics-department joint replacement operation system, the optimization matching speed of the to-be-operated-joint real skeleton local is high, and operation is easy.

The invention provides a bent axle box ventilation labyrinth with simple structure and good oil gas separating effect. The main body of the labyrinth is a cuboid cavity; the main body has an air outlet and an air inlet; multiple groups of diaphragms are arranged between the air inlet and the air outlet in the cavity at intervals; each group of diaphragms is composed of a first diaphragm close to the direction of the air inlet and a second diaphragm close to the direction of the air outlet; the distance of the first diaphragm and the second separation is 2-10mm; a first gas passing clearance is arranged between the top part of the first diaphragm and the top plate of the main body; a through hole is arranged at the center of the first diaphragm; a second gas passing clearance is arranged between the side of the diaphragm and the side plate of the main body; oil returning holes are arranged at the bottoms of the first diaphragm and the second diaphragm; the bent axle box ventilation labyrinth of the invention utilizes the clearances formed between the inner wall of the main body and the diaphragm and between the diaphragms as oil and gas flowing channels to lead oil and gas to flow in the narrow clearances, thus increasing the contact areas and crashing times between the oil and gas and the diaphragm as well as between the oil and gas and the inner wall of the main body, and improving the effect of oil and gas separation.