Patents
Literature

94results about How to "Informative" patented technology

Obstacle avoidance robot based on visual feature binding and reinforcement learning theory

The invention discloses an obstacle avoidance robot based on visual feature binding and a reinforcement learning theory. Visual feature binding and the reinforcement learning theory are utilized, and depth image information is integrated to learn priori knowledge of complex environment on the basis of a convolutional network and image matching technology; on the basis of a feature binding mechanism, a result is fed back to obstacle avoidance of the obstacle avoidance robot to make a decision intelligently, quickly, efficiently and accurately; an MNIST handwriting dataset is adopted, different control signal meanings are given to numbers 0-9 respectively, a Kinect sensor is utilized to collect RGB images with moving distance containing handwriting pictures in the complex environment, and an image matrix is created and output through gray-scale processing; image matching is performed, pictures after being matched are subjected to binarization processing and substituted into a well-trained image recognition model based on a convolutional neutral network, numbers in front of the sensor are judged finally, and corresponding control signals are transmitted to a robot control module through a host to realize intelligent obstacle avoidance.
Owner:CHINA UNIV OF MINING & TECH +2

Beef quality multi-parameter simultaneous detection method by multichannel near-infrared spectroscopy

The invention discloses a beef quality multi-parameter simultaneous detection method by multichannel near-infrared spectroscopy, belonging to the rapid nondestructive detection range of beef quality. The method comprises the following steps: first, arranging a spectral detection system comprising an 8-channel fiber multiplexer, two spectrometers with different wavelength ranges and probe units ina movable control cabinet, wherein the 8-channel fiber multiplexer is respectively connected with the spectrometers and the probe units; embedding in a production line technology to carry out on-linedetection to obtain optimum spectra data, wherein the quality detection of carcass and cut meat can be simultaneously satisfied; establishing a calibration prediction model, evaluating model effects by using relative analytical errors, carrying out model analysis on a large number of sample experimental data, respectively establishing prediction models of tenderness and water content to obtain a predicted value of each index, and then plugging the obtained predicted values into a beef multi-parameter comprehensive quality evaluation model which is established based on the detected indexes to obtain a final meat grade. According to the invention, the spectral information in the visible near-infrared band is collected, the information is abundant, the detection indexes have extensibility, and the method can be suitable for the requirement of production line detection.
Owner:CHINA AGRI UNIV

Dual-optical-frequency-comb linear spectral encoding imaging method

The invention discloses a dual-optical-frequency-comb linear spectral encoding imaging method. According to the method, two optical frequency combs with locked repetition frequencies are used, and thetwo optical frequency combs output signal light and local oscillator light respectively. The signal light is divided into detection light and reference light. The detection light passes a spectral encoder to be focused on the surface of a sample to be measured, the sample to be measured is encoded, and then the sample to be measured and the local oscillator light are subjected to dual-optical-frequency-comb heterodyne interference. The reference light directly interferes with the local oscillator light. Two interference pulse signals are generated and are converted into digital signals by signal acquisition. Optical reference compensation detection jitter is used, and a sample absolute distance with high precision and high stability, a linear reflectivity and phase information are obtained. The sample is moved linearly, continuous measurement is performed, and a 3D topographic image and a 2D reflectance image with high precision are achieved. The dual-optical-frequency-comb linear spectral encoding imaging has the advantages of simple structure, high measurement speed, high precision, a large amount of information and high expansion.
Owner:EAST CHINA NORMAL UNIVERSITY

Polarized SAR classification method based on deep learning of shallow-layer characteristics and T-matrix

ActiveCN107292317AInformativeThe classification result is accurateImage enhancementImage analysisClassification resultTest sample
The invention discloses a polarized SAR classification method based on deep learning of shallow-layer characteristics and T-matrix. The polarized SAR classification method based on deep learning of shallow-layer characteristics and T-matrix mainly solves the problem that the prior art is low in the classification correct rate for the same natural object having obvious difference on the scattering information and different natural objects having similar scattering information. The polarized SAR classification method based on deep learning of shallow-layer characteristics and T-matrix includes the following implementation steps: 1, filtering an original polarized SAR image; 2, extracting the polarized shallow-layer characteristics of the polarized SAR image after filtering; 3, fusing the shallow-layer characteristics with the polarized SAR data after filtering to construct training simples and a test samples; 4, learning the training samples through a convolution neural network; and 5, classifying the test samples by means of the convolution neural network which is obtained through learning, and obtaining the final polarized SAR natural object classification result. The polarized SAR classification method based on deep learning of shallow-layer characteristics and T-matrix is high in the classification correct rate of the polarized SAR target natural objects, has good experimental effect on classification of the natural object targets of a large area, and can be applied to target identification and natural object classification of a large scene.
Owner:XIDIAN UNIV

Image computer-aided diagnosis method for multi-sequence nuclear magnetic resonance images

The invention discloses an image computer-aided diagnosis method for multi-sequence nuclear magnetic resonance images, belongs to the field of computer-aided diagnosis, and relates to a computer-aided diagnosis method for multi-sequence image processing, texture feature extraction, classification and decision fusion of magnetic resonance imaging (MRI)-based T1WI, T2WI, an arterial phase, a portal vein phase and an equilibrium phase. According to the method, five sequences of the T1WI, the T2WI, the arterial phase, the portal vein phase and the equilibrium phase of MRI are integrated under a digital image processing and mode identification framework, and the image computer-aided diagnosis is completed by means of a neural network, a voting mechanism and a decision-making tree according to three levels of region of interest (ROI) processing, multi-sequence MRI classification and individual classification. By the method, multi-parameter, multi-sequence and multidirectional imaging is provided, and a combined classifier can select a sequence having the optimal distinguishing performance from the five sequences according to different stages of an anomaly structure to serve as the classification attribute of the corresponding stage. The image computer-aided diagnosis method has the advantages of rich information, clear levels and high classification accuracy.
Owner:DALIAN UNIV OF TECH

Point-of-interest information search device, system and method

ActiveCN102567385AInformativeSatisfied that the information is accurateSpecial data processing applicationsInformation deviceLibrary science
The invention discloses a point-of-interest information search device, system and method. The method includes: confirming searching conditions of the points of interest; obtaining benchmark information of the points of interest from a first information source according to the search conditions; obtaining pending-trial information of the points of information from a second information source according to the search conditions, wherein the benchmark information at least comprises accurate benchmark comparison information, and the pending-trial information at least comprises pending-trial comparison information and expending information which are identical in attributes; and judging whether the pending-trial comparison information is the same as benchmark comparison information, and outputting result information at least comprising the pending-trial comparison information and the expending information on yes judgment. The device comprises a confirming unit, a first search unit, a second search unit, a judging unit and an output unit. The system comprises a request end and the device. Simultaneously, requirements for abundant and accurate information amount of result information of points of interest search of users can be met.
Owner:SHANGHAI PATEO ELECTRONIC EQUIPMENT MANUFACTURING CO LTD

High-precision map-based visual loopback detection method and device

The embodiments of the present invention provide a high-precision map-based visual loopback detection method and device. According to the method and device, element information in a high-precision mapis matched with information extracted from a real-time image, and whether the element information in the high-precision map and the information extracted from the real-time image are information of the same location is judged; when it is detected that the element information in the high-precision map and the information extracted from the real-time image are the information of the same location,an error function is constructed through position coordinates in the high-precision map and positions in a real-time visual odometer; and a real-time pose is optimized by minimizing the error function, and therefore, the real-time pose in the visual odometer can be optimized, more accurate motion trajectories can be obtained, and the accuracy of the precise position calculation of the visual odometer can be improved. With the high-precision map-based visual loopback detection system adopted, more loopback detection constraints can be brought about; and since the information of the high-precision map is abundant, the accuracy of the high-precision map is higher than that of road sign points in a traditional visual odometer, and therefore, the accuracy of the real-time visual odometer can beimproved.
Owner:WUHAN ZHONGHAITING DATA TECH CO LTD

Gray image watermark embedding and extracting method based on singular value decomposition

The invention discloses a gray image watermark embedding and extracting method based on singular value decomposition. The gray image watermark embedding and extracting method based on singular value decomposition comprises the flowing steps: conducting singular value decomposition on an image and obtaining matrix components; conducting encoding and storage on singular values of a diagonal matrix of a singular value matrix and obtaining watermark signals; blocking a hosting image and choosing image blocks high in masking effect; conducting singular value decomposition on the image blocks and embedding the watermark signals into the image blocks; and conducting inverse transformation of the singular value decomposition on the image blocks. The gray image watermark embedding and extracting method based on singular value decomposition has the advantages that the embedding of the gray image watermark image is achieved, the watermark comprises a plurality of straight information, no using limiting exists on the size of the watermark image, and requirements of enterprises concealing personalized enterprise logos as watermarks can be met; the watermark embedding and extracting method can resist a plurality of image aggressive behaviors, and algorithm robustness is high; gray image watermark embedding and extracting method based on singular value decomposition can effectively detect watermark information content, solves the problem of image being illegally issued and the like, and is of great value for healthy and fast development of internet industries.
Owner:DALIAN NATIONALITIES UNIVERSITY

Visual detection system and method for through hole of transparent plate

The invention discloses a visual detection system and method for a through hole of a transparent plate . A combined light source is controlled by a controller to perform time-sharing switching of multiple lighting modes; at least one imaging device is controlled to carry out imaging scanning in different illumination modes; multi-channel image acquisition of shapes and defects of the through holeis realized; the information amount of image acquisition is greatly enriched; the accuracy and the precision of the through hole shape information and the defect information are improved; finally, theprocessor is used for analyzing based on a large amount of collected image data to obtain the shape precision and internal defects of the through hole in the transparent plate; the detection rate andthe recognition rate of the through hole defects are ensured; and a three-dimensional distribution diagram of the through hole defects in the transparent plate is accurately constructed, so that important reference indexes can be provided for guiding defect causes, adjusting a production process and grading material finished products on line, the real-time performance is high, the reaction speedis high, the online improvement of the processing quality is facilitated, and the production operation efficiency and the product yield are greatly improved.
Owner:湖南科创信息技术股份有限公司 +1
Who we serve
  • R&D Engineer
  • R&D Manager
  • IP Professional
Why Eureka
  • Industry Leading Data Capabilities
  • Powerful AI technology
  • Patent DNA Extraction
Social media
Try Eureka
PatSnap group products