118results about How to "Unclear solution" patented technology

The invention discloses a short text feature extraction method that performs feature extraction on a short text based on a knowledge base and a syntactic analysis method. The method comprises a model training process and a feature extraction process. The method comprises: performing training according to training set data; performing validation by using validation set data, and obtaining a weight set W that corresponds to a highest accuracy rate and a training model M that corresponds to the highest accuracy rate; after the feature extraction process performs processing for test set data, assigning the weight set W to each category; mapping the short text in a conceptual space by using an ESA algorithm, thereby obtaining an interpretation vector of the short text; and obtaining a topic vector through LDA, and using the vector as a final feature vector of the short text and a feature of the short text. The method provided by the invention can solve the problem that the short text is sparse in text feature and unclear in theme; and the method can reduce the difficulty in short text feature extraction processing, enhance the result of short text feature extraction, and improve accuracy of text classification.

The invention discloses an image acquisition method and device and an electronic device and belongs to the technical field of computers. The method includes the following steps: obtaining image information through a camera; detecting whether a jitter amplitude of the electronic device is in a preset range; and if the detection result shows that the jitter amplitude of the electronic device is in the preset range, acquiring the image information obtained by the camera. The device includes an obtaining module, a first detection module and an acquisition module. Through obtaining of image information in front of the camera, when it is detected that the jitter amplitude of the electronic device is in the preset range, the obtained image information is acquired; a problem that because of user hand vibration, obtained images are unclear is solved; and objectives that a shooting process of a user is simplified, the service life of the electronic device is improved and the acquired images are more clear are achieved.

The invention discloses a data synchronization method which comprises the steps that track process data from a handwriting device and an audio from a recording device are collected simultaneously, are synchronized according to a time attribute of the track process data and a timestamp of the audio and played. The data synchronization method collects the track process data and explains the audio simultaneously, and the track process data corresponds to the audio strictly according to the time attribute, so that a writing track and audio explanation are synchronized; the consistency of the audio explanation and frame display is improved greatly; and a content display effect is improved. The invention further discloses a device and a terminal which are used for achieving the method.

The invention discloses an automatic stool detector, comprising: a sample rack (1), a buffer area (2), an emergency position (3), a scanning part (4), a sample tray (5), a mechanical arm (6), a stirring Mechanism (7), sample loading unit (8), colloidal gold transfer mechanism (9), colloidal gold stack (10), microscope inspection unit (11), immunophotography unit (12), reaction chamber (13), colloidal gold waste material recovery (14). A mechanical arm is arranged above the sample tray. The robotic arm grabs the sample bottle from the sample rack to the station on the sample tray. The stirring mechanism stirs the material in the sample vial. The sample adding part adds the reagent into the sample bottle. The microscope inspection department conducts a microscope inspection on the materials in the sample bottle. The colloidal gold transfer mechanism extracts the colloidal gold from the colloidal gold stack (10), adds a sample, and then transfers the colloidal gold added with the sample to the reaction chamber (13). Immunophotographic department takes pictures to identify positive and negative. The colloidal gold waste recycling mechanism (14) recycles the colloidal gold.

The invention provides an anti-fog controlling method for a refrigerator camera. A refrigerator comprises a refrigerator body, a door body and a heating wire used for heating the camera. The anti-fog controlling method comprises the following steps that when the refrigerator starts to operate, the heating wire heats the camera with the power P1; when someone is detected to get close to the refrigerator and is about to open the refrigerator door body, the power of the heating wire is regulated and the heating wire heats the camera at the power P2; whether the refrigerator door body is open or not is detected; if the refrigerator door body is open, whether the refrigerator door body is closed or not is detected; if the refrigerator door body is closed, the power of the heating wire is regulated to recover to P1 to heat the camera; if the refrigerator door body is not closed, the camera temperature T1 and the environment temperature T2 are detected and compared; if T1 is greater than or equal to T2, the power of the heating wire is regulated to P1 to heat the camera; and if T1 is smaller than T2, the power of the heating wire is regulated to P3, and the camera is heated. The anti-fog controlling method for the refrigerator camera solves the problem that the shooting is vague and unclear due to the fogged camera, and can lower the energy consumption.

The invention discloses a foggy weather image reconstruction method based on polarization, comprising following steps: 1, obtaining the polarization images in different polarization angles; 2, calculating the Stokes vector value of an object image; 3, estimating the atmosphere light intensity value of the infinite distance by using light channel priori rules; 4, estimating the atmosphere light polarization degree by using a filtering method; 5, reconstructing a fogless scenery image; 6, carrying out DOP partial index enhancement to the fogless image. According to the invention, the fogless image from the foggy weather image can be reconstructed well; and the definition of the image is improved.

The invention discloses a network video browsing method, which comprises the following steps that: at least one video acquirer acquires a video from a video server; a video encoder converts each frame of the video acquired by at least one video acquirer into a picture according to a tine sequence; and the video encoder sends the converted pictures to a client in real time. According to the network video browsing method, the problem of unclear video pictures caused by easy frame loss in the network video browsing method in the prior art, is solved; each frame of the video is converted into the picture according to the time sequence, and the video is transmitted to the client in the form of pictures, so that the phenomenon of frame loss during video browsing is avoided; and furthermore, the resolution of the pictures can be adjusted, so that the bandwidth occupation rate can be controlled.

The invention provides a GABOR characteristic based face sketch synthetic method and system. The synthetic method comprises the following steps: dividing a face photo to be synthesized, a face photo sample and a face sketch sample into mutually overlapped pixel blocks; extracting the Gabor characteristics of each pixel block, and acquiring the Stein divergence matrix on the basis of the covariance matrix of the Gabor characteristics of the pixel block; acquiring the optimal weight on the basis of the Stein divergence matrix and the rebuilding coefficient of each pixel block of the face photo sample; adopting the optimal weight for weighting synthesis on the pixel block at the corresponding position of the face sketch sample to acquire a synthetic face sketch pixel block; integrating the synthetic face sketch pixel block to acquire the synthetic face sketch corresponding to the face photo to be synthesized. The synthetic method and system can synthesize the face sketch in higher quality, and can be applied to criminal investigation.

The invention discloses a shooting method of an object. The shooting method of the object comprises the following steps of: after shooting is started, collecting an object image within a preset time according to a preset focal distance; performing image combination based on all the acquired object images so as to generate a combined image corresponding to the object. The invention also discloses a mobile terminal. As the object images are acquired at different focal distances, and all the object images are combined to generate the combined image corresponding to the object, the problem that the combined object image is unclear is solved, and a clear combined image of the object can be taken on exactly for a user.

The invention provides an image preprocessing method and an image preprocessing system for parking spot detection. The method comprises the following steps: acquiring a parking spot scene image, wherein the reflective part of a parking spot scene is displayed in the parking spot scene image; and processing the parking spot scene image using a defogging algorithm, thus highlighting the information about the parking spots in the parking spot scene image and reducing the influence of the reflective part on parking spot detection. Problems, for example, light reflection of a site affects parking spot line extraction, parking spots are not clear in a dark environment, and an image has many noisy points due to adverse factors of a camera, can be solved effectively. The efficiency and accuracy of parking spot detection in an image are improved.

The invention provides an artificial gas cap and edge water two-way drive development reservoir screening and evaluating method for a fault block oil reservoir. The method comprises the following steps: (1), developing fine strata comparison and structural interpretation of a candidate fault block oil reservoir, establishing a reservoir geologic model, and determining a reservoir parameter; (2), establishing a reservoir numerical model, developing development history fitting, and correcting the reservoir geologic model and the reservoir parameter; (3), analyzing the potential of increasing the reservoir recovery ratio through two-way drive, and demonstrating the necessity of developing two-way drive development; and (4), judging whether the candidate fault block oil reservoir is suitable for two-way drive development or not according to the corrected reservoir parameter. By means of the artificial gas cap and edge water two-way drive development reservoir screening and evaluating method for the fault block oil reservoir, the problem that a screening and potential-evaluating method of a two-way drive development unit of a fault block oil reservoir having extra-high water-cut stage is undefined and nonstandard can be solved; and thus, the technical application effect of increasing the recovery ratio by developing two-way drive in the last period of development of the fault block oil reservoir and the development adjustment plan compiling level can be effectively increased.

The invention discloses an infrared image compression method and an infrared image compression system applied to a contact network; the method comprises the following steps: S1, obtaining an infrared temperature image stream; S2, successively taking out each frame infrared temperature image in the infrared temperature image stream; S3, judging whether the current frame infrared temperature image is a first frame image of the infrared temperature image stream; S4, creating a blank infrared video compressed file and a corresponding blank index file, and writing a file header in the index file; S5, carrying out compression processing as well as frame information generation and writing on the infrared temperature image; and S6, judging whether the current frame infrared temperature image is a last frame image of the infrared temperature image stream. According to the infrared image compression method and the infrared image compression system applied to the contact network provided by the invention, through carrying out the compression processing on each frame infrared temperature image in the infrared temperature image stream collected by an infrared camera, the space for storing the infrared temperature image stream is reduced.

The invention discloses an icing detection method and system for inspecting an overhead transmission line based on an unmanned aerial vehicle. The method comprises the steps of obtaining a video shotin the process of inspecting the overhead transmission line by the unmanned aerial vehicle; extracting a key frame image from the video; processing the key frame image to respectively obtain a gray image, a contour image and a texture image; extracting gray features of the gray image, contour features of the contour image and texture features of the texture image; carrying out normalization on theextracted gray features, contour features and texture features, performing series fusion on the gray features, contour features and texture features after normalization processing, and acquiring fusion features; and inputting the fusion features into a pre-trained convolutional neural network model, and outputting a classification result about whether icing occurs or not.

A polarized light endoscope device comprises an optical prism, an RGB image sensor, a polarized light image sensor and an image processor, the optical prism divides one light beam output by an optical lens into two light beams, one light beam is output to the RGB image sensor, and the other light beam is output to the polarized light image sensor; the RGB image sensor collects RGB optical signals and converts the RGB optical signals into RGB pixel data; the polarized light image sensor collects polarized light signals and converts the polarized light signals into polarization information pixel data; and the image processor synchronously performs image processing on the RGB pixel data and sharpening processing on the polarization information pixel data, and outputs and displays the obtained RGB image and the polarization information image, so that when blood, tissue crumbs, fog and the like appear in an operation environment, the problem that the output image of the endoscope device is not clear is solved through sharpening processing on the polarization information pixel data.

The invention discloses an electronic document flowing method, which comprises the following steps: a user uploads an electronic document; a receiving party logs in a server, receives the electronic document through a document receiving module, and processes and archives the electronic document, and a corresponding identification is marked on the processed and archived electronic document through a marking component; according to the identification, an identification recognition module compares stage information preset in the identification recognition module with the corresponding identification to obtain a stage where the identification belongs to, and sends a processing stage where the electronic document is positioned to a document query module; a forwarding module inquires whether the receiving party needs to forward the electronic document or not, and the receiver selects whether the electronic document needs to be forwarded or not; and the user queries the processing stage where the electronic document is positioned through the document query module. The invention simultaneously discloses an electronic document flowing system. The user can clearly and accurately obtain the stage of the electronic document in the flowing process so as to effectively solve the problem that the electronic document flowing process is unclear in the prior art.

The invention belongs to the technical field of multimedia playing, which discloses multimedia playing equipment and a method and device for displaying texts in a file of the multimedia playing equipment. The method comprises the following steps of: acquiring bitmap data of fonts to be played; acquiring colors of font outlines according to colors corresponding to the bitmap data, wherein the colors of the font outlines are different from the colors corresponding to the bitmap data; drawing the font outlines of the fonts to be displayed according to the colors of the font outlines; drawing the fonts to be displayed according to colors corresponding to the bitmap data; and displaying the drawn fonts. According to the invention, the problem that external store subtitle is indistinct because a main color of a video is similar or the same as the display color of the external store subtitle is better solved; and the great convenience for watch of a user is realized.

The invention discloses an image self-learning enhancement method and system. The method comprises the following steps that an image self-learning assessment unit is trained, and optimal enhancement algorithms corresponding to the types of different images are determined according to the types of different images; and a real-time image acquired by an image acquisition unit is transmitted to a real-time image enhancement unit, and the real-time image enhancement unit acquires image feature parameters of the real-time image acquired at present, determines the type of the image according to the real-time image feature parameters, acquires the optimal enhancement algorithm corresponding to the image type from the image self-learning assessment unit and conducts enhancement processing on the real-time image by using the optimal enhancement algorithm corresponding to the image type. Accordingly, the problem that the sea area images acquired in real time are not clear can be effectively solved, and the method and system are beneficial for technicians to observe the maritime environment and have a wide prospect in the fields of national defense, security protection, maritime search and rescue and the like.

The invention discloses a license plate information collecting device and method. The license plate information collecting device comprises a processing module and at least two camera modules and is characterized in that each camera module is arranged at a preset angle and can take license plate pictures from different angles; the processing module is used for performing quality judgment on the multiple license plate pictures and selecting the license plate pictures satisfying preset identification requirements. By the license plate information collecting device, the technical problem that dueto the fact that an existing license plate camera only uses one camera module to take the pictures in a fixed range from a fixed angle, the acquired license plate pictures are unclear and incompleteand cannot be identified by a license plate identification system is solved.

The invention discloses a mobile terminal and a shooting method thereof. The shooting method includes that after beginning shooting, gathering object images according to a preset time interval and a corresponding preset focal length; selecting all the object images gathered in the preset time length or selecting a preset number of object images; carrying out additive operation on the selected object images to generate a composite image corresponding to the object movement track. Through gathering the corresponding object images of an object (such as thunder) through zooming treatment and compositing all the object images to generate the corresponding composite image, the shooting method solves the problem of unclear composite image and improves the image gathering definition, and accordingly the clear composite image corresponding to the object can be precisely presented to a user.

The invention relates to a vehicle-mounted traffic rule and regulation violation intelligent monitoring apparatus. The vehicle-mounted traffic rule and regulation violation intelligent monitoring apparatus comprises a power source management module, an image acquisition module, a storage module, a control module and an information transmission module. The vehicle-mounted traffic rule and regulation violation intelligent monitoring apparatus is characterized in that the monitoring apparatus further comprises a triggering module and a map module. When traffic rule and regulation violation occurs, a sensor controller in the triggering module feeds back and transmits received triggering signals emitted by a sensor to an application processor; the application processor edits pictures taken by a camera 10 seconds before and after a time point when the application processor receives the triggering signals and accurate vehicle traffic rule and regulation violation location information acquired by the map module into a picture stream file of which the total time is 20 seconds; a wireless module uploads the picture stream file to a law enforcement agency server through 3G or / 4G or / WIFI; and an evident obtaining process of traffic rule and regulation breaking is completed.

The invention discloses an image display layout method, device and equipment and a storage medium, and relates to the field of computer vision and deep learning. The specific implementation scheme comprises the steps determining a plurality of display areas of a display page and the size of each display area in the plurality of display areas according to a predetermined rule; determining a scalingratio of each to-be-displayed image in all the to-be-displayed images based on the sizes of the display areas, and adjusting the determined number of the display areas and the sizes of the display areas when the scaling ratio exceeds a threshold value; and displaying the to-be-displayed image in the display area based on the adjusted number of the display areas and the adjusted sizes of the display areas. By detecting and correspondingly adjusting the scaling of each to-be-displayed image, the scaling of each to-be-displayed image is smaller than the threshold value, the problem that the image in the image layout is not clear after being amplified is solved, and the image viewing experience of a user is improved.

The invention relates to an image processing method and device and a terminal. The method comprises the steps that when amplification operation performed on an image is detected, characters in the image are identified and the amplification proportion is determined; the amplification image of the image is drawn based on the amplification proportion and the characters; and the amplification image is displayed. With application of the image processing method and device and the terminal, the terminal can draw the amplification image again based on the amplification proportion of amplification operation and the characters in the image when a user amplifies the image so that definition of the characters in the amplification image can be guaranteed, the problems that the characters are blurry and unclear after amplification of certain images can be solved and user experience can be optimized.

The invention discloses a method and system for enabling a terminal to play DVB services. The method comprises the steps that a request of a playing channel is received through a DVB server, and the type of the playing channel is judged; if the playing channel is an audio and video program, audio and video data corresponding to the playing channel are filtered, demultiplexed and sectionally multiplexed and sent to the terminal to be played; if the playing channel is an image and text program, image and text data corresponding to the image and text program are downloaded, the home page address corresponding to the image and text program is sent to the terminal, and the terminal plays the image and text data according to the home page address. Thus, the terminal can play the DVB services such as television programs, image and text programs and webpages, is high in playing definition, has no delay, cannot be affected by telecommunication networks and brings great convenience.

The invention discloses a method for extracting a multi-prime interface based on point collective elements, which comprises the following steps: step 1, acquiring point cloud data of an image, and carrying out Gaussian filter denoising; step 2, converting a point set based on a spherical neighborhood, and highlighting a demarcation point; step 3, calculating the minimum circumscribed ball parameter, obtaining the minimum circumscribed ball point set, and utilizing the spherical point to shrink to obtain a single interface; step 4, calculating the minimum circumferential ball parameters again,obtaining the minimum circumferential ball point set, and dividing the multi-partition interfaces by utilizing the spherical point shrinkage sequence. A Gaussian filter is used for denoise in that multi-quality interface extraction method of the invention, so the correctness of the model in the processing process is ensured and the influence of the noise point is reduced. Through the weighted boundary point highlighting method, the purpose of reasonable use of neighborhood information is better ensured, the boundary point is accurately demarcated, the problem of unclear interface caused by neighborhood influence is solved, and the veracity of boundary point judgment is better ensured in the model processing.

The invention discloses a method and device for supporting multiple persons to capture a projection screen playing picture and a computer storage medium, and the method comprises the following steps: an intelligent terminal establishes communication connection with an intelligent display terminal, and carries out projection screen playing; the preset number of third-party intelligent terminals establish communication connection with the intelligent display terminal in a preset mode; screen capture requests of a preset number of third-party intelligent terminals and timestamps corresponding to the screen capture requests are obtained; according to the preset number of screen capture requests of the third-party intelligent terminals and the timestamps corresponding to the screen capture requests, a playing picture corresponding to a playing file of the intelligent display terminal is obtained; anda playing picture corresponding to the intelligent display end playing file is sent to a preset number of third-party intelligent terminals. The problem that in the prior art, when a speaker carries out screen projection display through an intelligent terminal and intelligent display equipment, a viewer can only record the screen projection content in the mode that a third-party intelligent terminal takes a picture, and consequently the taken picture is not clear is solved.

An embodiment of the invention provides a jitter correcting method and terminal equipment, which relate to the technical field of terminals, and are used for solving the problem that a captured videoor picture is unclear caused by jitter in the shooting process. The jitter correcting method comprises the steps of: acquiring preview image data in real time; determining a first correction strategyaccording to a jitter type of the preview image data; and correcting a captured image according to the first correction strategy. The jitter correcting method and the terminal equipment are used for shooting pictures or videos.

The invention provides an image processing method and device, and a three-dimensional scanning system. The method comprises the following steps: acquiring at least two groups of two-dimensional images obtained by shooting a target object by adopting at least two different exposure parameters; screening a plurality of images with normal exposure from the at least two groups of two-dimensional images; and performing three-dimensional reconstruction based on the connected regions of the plurality of images with normal exposure, and generating a three-dimensional image of the target object. A problem that the precision of the impressed three-dimensional data is affected due to unclear shot pictures caused by a single exposure parameter is solved.

The invention relates to a temporary parking space management and control device, system and method. The device comprises a vehicle detection unit, a roadside unit, a parking space information indication unit, a voice prompt unit and an information display unit; the vehicle detection unit is used for detecting the information of vehicles near and in a temporary parking space, and transmitting the information to the roadside unit; after the information is processed by the roadside unit, the parking space information indication unit, the voice prompt unit and the information display unit make corresponding responses, and the obtained parking information is sent to a server, and then the server sends the parking information to a mobile terminal. Each unit of the system can timely inform a driver of the condition that temporary parking is allowed near the temporary parking space through voice, light-emitting contour and character display, can send parking information can to the mobile terminal, so that closed-loop prompt information is formed, the driver can check the remaining parking time at any time, the problem that traffic is affected by overtime parking due to the fact that parking time is forgotten or parking of a car owner in a fixed parking space is solved.