Patents by Inventor Sen Jia
Sen Jia has filed for patents to protect the following inventions. This listing includes patent applications that are pending as well as patents that have already been granted by the United States Patent and Trademark Office (USPTO).
-
Publication number: 20230377190Abstract: A method for training models is provided. The method includes: inputting an image training sample corresponding to a current iteration into a current posture detection network model, and acquiring a first loss function corresponding to the current iteration; re-projecting the current output result of the current posture detection network model, and acquiring a second loss function corresponding to the current iteration; and acquiring a posture detection network model for a next iteration by performing backpropagation on the current posture detection network model, and achieving training of the posture detection network model by performing the next iteration before an iteration end condition is met.Type: ApplicationFiled: October 26, 2021Publication date: November 23, 2023Inventor: Sen JIA
-
Patent number: 11586863Abstract: Provided are an image fusion classification method and device. The method includes that: a three-dimensional weight matrix of a hyperspectral image is obtained by use of a Support Vector Machine (SVM) classifier (101); superpixel segmentation is performed on the hyperspectral image to obtain K superpixel images, K being a positive integer (102); the three-dimensional weight matrix is regularized by use of a superpixel-image-based segmentation method to obtain a regular matrix (103); and a class that a sample belongs to is determined according to the regular matrix (104).Type: GrantFiled: March 22, 2021Date of Patent: February 21, 2023Assignee: SHENZHEN UNIVERSITYInventors: Sen Jia, Bin Deng, Jiasong Zhu, Lin Deng, Qingquan Li
-
Imaging method and device for nonlinear parallel magnetic resonance image reconstruction, and medium
Patent number: 11579229Abstract: There are provided a parallel rapid imaging method and device based on a complex number conjugate symmetry of multi-channel coil data and nonlinear GRAPPA image reconstruction, and a medium. The imaging method includes: obtaining virtual conjugate coil data by expanding the actual multi-channel coil data; combining actual multi-channel coil data and virtual multi-channel coil data to obtain a linear data term and a nonlinear data term; calibrating weighting factors of the linear data term and the nonlinear data term by using combined low-frequency full-sampling data (margins of the low-frequency full-sampling data includes parts of high-frequency data); reconstructing data which is under-sampled in a high-frequency region according to the calibrated weighting factors; fusing the low-frequency full-sampling data and the reconstructed data for the high-frequency region.Type: GrantFiled: July 24, 2018Date of Patent: February 14, 2023Assignee: SHENZHEN INSTITUTES OF ADVANCED TECHNOLOGYInventors: Haifeng Wang, Dong Liang, Sen Jia, Xin Liu, Hairong Zheng -
IMAGING METHOD AND DEVICE FOR NONLINEAR PARALLEL MAGNETIC RESONANCE IMAGE RECONSTRUCTION, AND MEDIUM
Publication number: 20210356545Abstract: There are provided a parallel rapid imaging method and device based on a complex number conjugate symmetry of multi-channel coil data and nonlinear GRAPPA image reconstruction, and a medium. The imaging method includes: obtaining virtual conjugate coil data by expanding the actual multi-channel coil data; combining actual multi-channel coil data and virtual multi-channel coil data to obtain a linear data term and a nonlinear data term; calibrating weighting factors of the linear data term and the nonlinear data term by using combined low-frequency full-sampling data (margins of the low-frequency full-sampling data includes parts of high-frequency data); reconstructing data which is under-sampled in a high-frequency region according to the calibrated weighting factors; fusing the low-frequency full-sampling data and the reconstructed data for the high-frequency region.Type: ApplicationFiled: July 24, 2018Publication date: November 18, 2021Inventors: Haifeng WANG, Dong LIANG, Sen JIA, Xin LIU, Hairong ZHENG -
Publication number: 20210209426Abstract: Provided are an image fusion classification method and device. The method includes that: a three-dimensional weight matrix of a hyperspectral image is obtained by use of a Support Vector Machine (SVM) classifier (101); superpixel segmentation is performed on the hyperspectral image to obtain K superpixel images, K being a positive integer (102); the three-dimensional weight matrix is regularized by use of a superpixel-image-based segmentation method to obtain a regular matrix (103); and a class that a sample belongs to is determined according to the regular matrix (104).Type: ApplicationFiled: March 22, 2021Publication date: July 8, 2021Inventors: Sen JIA, Bin DENG, Jiasong ZHU, Lin DENG, Qingquan LI
-
Patent number: 10783371Abstract: The present invention provides a Gabor cube feature selection-based classification method for hyperspectral remote sensing images, comprising the following steps: generating three-dimensional Gabor filters according to set frequency and direction parameter values; convoluting hyperspectral remote sensing images with the three-dimensional Gabor filters to obtain three-dimensional Gabor features; selecting three-dimensional Gabor features, classification contribution degrees to various classes of which meet preset requirements, from the three-dimensional Gabor features; and classifying the hyperspectral remote sensing images by a multi-task joint sparse representation-based classification means by using the selected three-dimensional Gabor features. The present invention is based on the three-dimensional Gabor features, and the used three-dimensional Gabor features contain rich local change information of a signal and are competent in feature characterizing.Type: GrantFiled: May 15, 2018Date of Patent: September 22, 2020Assignee: SHENZHEN UNIVERSITYInventors: Sen Jia, Jie Hu, Yao Xie, Linlin Shen
-
Patent number: 10509984Abstract: The present invention provides a method for extracting and classifying features of hyperspectral remote sensing image, including: an sampling step, a binarizing step, a coding step, a statistical calculating step, a concatenating step, and a classifying step. The present invention further provides a system for extracting and classifying features of hyperspectral remote sensing image. The technical solution provided by the present invention can make full use of the contextual relationship between the spectral domain and the spatial domain in a hyperspectral remote sensing image by extending two-dimensional LBPs into three-dimensional LBPs, and has good robustness to noise by introducing a relaxation threshold discrimination operation.Type: GrantFiled: May 13, 2018Date of Patent: December 17, 2019Assignee: SHENZHEN UNIVERSITYInventors: Sen Jia, Jie Hu, Lin Deng
-
Publication number: 20180268195Abstract: The present invention provides a Gabor cube feature selection-based classification method for hyperspectral remote sensing images, comprising the following steps: generating three-dimensional Gabor filters according to set frequency and direction parameter values; convoluting hyperspectral remote sensing images with the three-dimensional Gabor filters to obtain three-dimensional Gabor features; selecting three-dimensional Gabor features, classification contribution degrees to various classes of which meet preset requirements, from the three-dimensional Gabor features; and classifying the hyperspectral remote sensing images by a multi-task joint sparse representation-based classification means by using the selected three-dimensional Gabor features. The present invention is based on the three-dimensional Gabor features, and the used three-dimensional Gabor features contain rich local change information of a signal and are competent in feature characterizing.Type: ApplicationFiled: May 15, 2018Publication date: September 20, 2018Inventors: Sen Jia, Jie Hu, Yao Xie, Linlin Shen
-
Publication number: 20180260657Abstract: The present invention provides a method for extracting and classifying features of hyperspectral remote sensing image, including: an sampling step, a binarizing step, a coding step, a statistical calculating step, a concatenating step, and a classifying step. The present invention further provides a system for extracting and classifying features of hyperspectral remote sensing image. The technical solution provided by the present invention can make full use of the contextual relationship between the spectral domain and the spatial domain in a hyperspectral remote sensing image by extending two-dimensional LBPs into three-dimensional LBPs, and has good robustness to noise by introducing a relaxation threshold discrimination operation.Type: ApplicationFiled: May 13, 2018Publication date: September 13, 2018Inventors: Sen Jia, Jie Hu, Lin Deng
-
Publication number: 20160224702Abstract: A method of calculating voltage and power of large-scaled photovoltaic power plant includes following steps. Environment models of varies locations within the photovoltaic power plant are obtained. A photovoltaic display model is established by establishing a photovoltaic cell model, combining the photovoltaic cell model with the environment models, determining a combination of the environmental data on photovoltaic panels and the photovoltaic cell model, and determining a quantitative relationship between a photovoltaic cell power generation state and photovoltaic environment. An inverter model is obtained by modeling inverters connected to photovoltaic cells. A grid-side model is obtained. An electrical energy and voltage forecast model is constructed by integrated the photovoltaic display model, the inverter model, and the grid-side model.Type: ApplicationFiled: January 31, 2016Publication date: August 4, 2016Inventors: MING MA, NING-BO WANG, SHI-EN HE, YAN-HONG MA, XU-SHAN HAN, ZI-FEN HAN, HUAI-SEN JIA, PENG ZHANG, GUANG-TU LIU, LONG ZHAO, QIANG ZHOU, DING-MEI WANG, JIAN-MEI ZHANG, QING-QUAN LV, MING-SONG WANG, ZHAO CHEN, YAN-LI ZHANG, KUN DING, JIN LI, SHI-YUAN ZHOU, LIANG LU, RONG HUANG, JIN-PING ZHANG, NIAN-ZONG BAI
-
Publication number: 20160025071Abstract: A method of computing theoretical power of wind farm based on sample wind turbine method. Historical real-time measured output data of sample wind turbines in the wind farm is statistically analyzed. A theoretical output of the sample wind turbines is analyzed based on a wind power output characteristic and a theoretical output design curve of wind turbine. A theoretical power calculation model of wind farm is constructed based on the historical real-time measured output data of sample wind turbines. A number of wind turbines actually running in real-time is obtained based on real-time information data of wind turbines in the wind farm. An actual operation data of the sample wind turbines is confirmed. The actual operation data is input into the theoretical power calculation model of wind farm. Real-time data of theoretical power of wind farm is output, and an actual operation result of wind farm is comparatively analyzed.Type: ApplicationFiled: July 27, 2015Publication date: January 28, 2016Inventors: NING-BO WANG, MING MA, QING-QUAN LV, YAN-HONG MA, GUANG-TU LIU, HUAI-SEN JIA, LONG ZHAO, ZI-FEN HAN, QIANG ZHOU, DING-MEI WANG, LIANG LU, JIAN-MEI ZHANG
-
Publication number: 20160025070Abstract: A method of calculating theoretical power of wind farm based on extrapolation of anemometer tower data includes following steps. A number of anemometer towers in a wind farm is selected, and analyzing historical data acquired by the number of anemometer towers. An air density is calculated based on the historical data of the number of anemometer towers. A power curve is calibrated based on the historical data of the number of anemometer towers. The power curve is fit based on a wind speed and a wind power of a fan head of each wind turbine based on the historical data. An theoretical power calculation extrapolation model of anemometer tower data is constructed. Real-time anemometer tower wind data and a calibrated air density are inputted into the theoretical power calculation model and calculating the wind power. The theoretical power is obtained.Type: ApplicationFiled: July 27, 2015Publication date: January 28, 2016Inventors: NING-BO WANG, MING MA, QING-QUAN LV, YAN-HONG MA, GUANG-TU LIU, HUAI-SEN JIA, LONG ZHAO, ZI-FEN HAN, QIANG ZHOU, DING-MEI WANG, LIANG LU, JIAN-MEI ZHANG
-
Publication number: 20150302313Abstract: A method of predicating ultra-short-term wind power based on self-learning composite data source includes following steps. Model parameters of an autoregression moving average model are obtained by inputting data. A predication result is obtained by inputting data required by wind power predication into the autoregression moving average model. A post-evaluation is performed to the predication result by analyzing error between the predication result and measured values, and performing model order determination and model parameters estimation again while the error is greater than an allowable maximum error.Type: ApplicationFiled: April 9, 2015Publication date: October 22, 2015Inventors: NING-BO WANG, LIANG LU, XU-SHAN HAN, ZI-FEN HAN, HUAI-SEN JIA, XIAO-YONG WANG, RONG HUANG, JIN-PING ZHANG
-
Patent number: 7804257Abstract: A control circuit (100) being adapted for identifying an adaptor of an electronic device includes an adaptor-side circuit (110) that is incorporated in the adaptor and a system-side circuit (120) that is incorporated in the electronic device. The adaptor-side circuit converts a high alternating voltage coming from a power source to a low direct voltage and has a positive output terminal and a negative output terminal. A voltage dividing circuit (30) is connected with the positive output terminal, and has a control terminal outputting a voltage. The system-side circuit includes a switching circuit (40) connected to the positive output terminal and the control terminal. The voltage outputted by the control terminal is used to switch on the switching circuit in order to energize the electronic device via the direct current voltage outputted by the positive and negative output terminals.Type: GrantFiled: December 1, 2006Date of Patent: September 28, 2010Assignees: Hong Fu Jin Precision Industry (ShenZhen) Co., Ltd., Hon Hai Precision Industry Co., Ltd.Inventor: Yu-Sen Jia
-
Publication number: 20080129123Abstract: A control circuit (100) being adapted for identifying an adaptor of an electronic device includes an adaptor-side circuit (110) that is incorporated in the adaptor and a system-side circuit (120) that is incorporated in the electronic device. The adaptor-side circuit converts a high alternating voltage coming from a power source to a low direct voltage and has a positive output terminal and a negative output terminal. A voltage dividing circuit (30) is connected with the positive output terminal, and has a control terminal outputting a voltage. The system-side circuit includes a switching circuit (40) connected to the positive output terminal and the control terminal. The voltage outputted by the control terminal is used to switch on the switching circuit in order to energize the electronic device via the direct current voltage outputted by the positive and negative output terminals.Type: ApplicationFiled: December 1, 2006Publication date: June 5, 2008Applicant: HON HAI PRECISION INDUSTRY CO., LTD.Inventor: YU-SEN JIA
-
Publication number: 20070008084Abstract: An electronic license plate system capable of changing information displayed thereon and method for the same is a display device installed on in a motor vehicle showing information according to random instructions given by a traffic control system, serving as an illuminated license plate for easy visual identification. The displayed contents of the license plate varies with time and location to prevent counterfeiting or a shut-down event, and alertness and road security may be raised in areas with poor visibility due to the illuminated license plate. The electronic license plate working with the traffic control system may reach the functions of unmanned charge, dynamic geographic information, unmanned traffic control, road use information, and traffic statistics, and may provide a functional intelligent traffic system (ITS) for prevention of license plate counterfeiting and vehicle theft.Type: ApplicationFiled: July 5, 2005Publication date: January 11, 2007Inventors: Sen-Jia Wu, Hao-Lan Tang, Yu-Tsung Hsu, Yu-Tang Huang
-
Publication number: 20040007917Abstract: The present invention provides a cascaded network power control system comprising at least a controlled power site and a control center. The controlled power site has a front end processor (FEP). The FEP at least has two network ports (an uplink network port and a downlink network port). The control center uses an optical fiber to connect the network ports of the FEP, and uses the network ports of the FEP to establish contact with another controlled power site. The network ports are connected with network ports of a preceding controlled power site and the next controlled power site for transmission of a normal/abnormal signal of an intelligent electricity device (IED) of the preceding controlled power site. The normal/abnormal signal is transmitted to the IED of the local controlled power site via an RS232 or RS485 port.Type: ApplicationFiled: July 15, 2002Publication date: January 15, 2004Applicant: ALLIS ELECTRIC CO., LTD.Inventors: Sen Jia Wu, Wen-Fu Tsai, Jiann-Yi Fann, Fu-Hsiu Chiang