公开(公告)号：US20180082423A1

公开(公告)日：2018-03-22

申请号：US15598686

申请日：2017-05-18

Applicant: SICHUAN UNIVERSITY

Inventor： Junfeng WANG ,  Peng TANG ,  Yulin JI ,  Zongan LIANG ,  Yihua DU

IPC: G06T7/00 ,  G06T7/13 ,  A61B5/00 ,  A61B6/00

CPC classification number: G06T7/0014 ,  A61B5/7225 ,  A61B6/50 ,  A61B6/5211 ,  A61B6/5217 ,  G06T7/13 ,  G06T7/74 ,  G06T2207/10116 ,  G06T2207/20016 ,  G06T2207/20081 ,  G06T2207/30061

Abstract: A lung lobe contour extraction method aiming at DR radiography, which includes the steps: Obtain the representative lung lobe contour template through offline training; initialize the extraction system of contour lobe area in the chest DR radiography; According to dimension of DICOM image, conduct three-layer pyramid decomposition on the images; Use Gabor filter bank to rebuild pending images; Convert residual error of the rebuilt image by the Gabor filter into black and white image; Use Zhan-Suen refinement algorithm to refine the black and white image; Call every template of the offline training and use them as the convolution kernel operator, conduct convolution on the contour imagep; Filter the local optimum convolution value with optimum possibility from the convolution results and combine the evaluations; Generate the lung lobe contour shape in accordance with the upper and lower templates of optimum matching and its position.

公开(公告)号：US20170346750A1

公开(公告)日：2017-11-30

申请号：US15602433

申请日：2017-05-23

Applicant: SICHUAN UNIVERSITY

Inventor： Junfeng WANG ,  Lixiang LIU ,  Fuchun SUN ,  Dong LIU ,  Shiping YANG

IPC: H04L12/807 ,  H04L12/801 ,  H04L1/16

CPC classification number: H04L47/27 ,  H04L1/0019 ,  H04L1/1607 ,  H04L1/1825 ,  H04L47/12

Abstract: A kind of self-adaptive network congestion control method based on SCPS-TP, which includes the following steps: The SCPS-TP's gateway source-end receives and transmits the packets to destination end; Judge if there is new packet received in accordance with the analyzed ACK; If there is no new packet received, when the duplicate ACK counter increase to a certain value, change the window size's growth pattern to linear self-adaptive pattern; If there is new packet received, the congestion control is in the exponential growth pattern. After window is enlarged, Diff is bigger than the set threshold value and the congestion control method is changed to linear self-adaptive pattern; If congestion control is in the linear self-adaptive pattern, adjust window size in accordance with Diff; The SCPS-TP's gateway source-end sends the packets in the packet loss buffer to destination end and sends new packets in accordance with the size of congestion window.

公开(公告)号：US20190191130A1

公开(公告)日：2019-06-20

申请号：US16219844

申请日：2018-12-13

Applicant: SICHUAN UNIVERSITY

Inventor： Junfeng WANG ,  Qin LUO ,  Ying LIU ,  Hao NIU ,  Xiaohui LI

IPC: H04N7/20 ,  H04B17/391 ,  H04N21/647 ,  H04N21/643 ,  H04L25/03

Abstract: This invention discloses a kind of reliable data transmission method based on hidden Markov model that specific to video transmission quality enhancement over lossy satellite links. It uses HMM model to forecast retransmission rate and adopts the network packet loss rate to reflect the network performance. The underlying network packet loss rate is used as the hidden state of HMM model and the retransmission rate is used as the observation state. By analyzing the historically transmitted data sessions, APRT establishes the relationship between the hidden state transition probability and the emission probability. Then, a robust and efficient retransmission rate prediction is obtained. This invention takes into full consideration of high BER in satellite network channels and adjusts the reliable level according to the satellite network channel characteristics so as to make tradeoff between instantaneity and quality. Consequently, it guarantee the QoS and QoE of video stream.

公开(公告)号：US20170353386A1

公开(公告)日：2017-12-07

申请号：US15602703

申请日：2017-05-23

Applicant: SICHUAN UNIVERSITY

Inventor： Junfeng WANG ,  Fuchun SUN ,  Lixiang LIU ,  Dong LIU ,  Shiping YANG

IPC: H04L12/801 ,  H04L12/807 ,  H04L12/26

CPC classification number: H04L47/12 ,  H04L43/0864 ,  H04L47/27

Abstract: A kind of congestion improvement method based on the QUIC protocol adds the information of round trip delay in the congestion algorithm, self-adaptive changes the value of α to judge the situation of current network through comparison between the RTT of last time and the current RTT and then adjusts the current target window value in accordance with the current network situation, changing the congestion window based on the cubic growth curve of the cubic algorithm. This improvement method can make the QUIC protocol judge the current network situation more timely and accurately and can make the congestion window change quickly to fully utilize the bandwidth. The maximum congestion window limitation 200 exists in the QUIC protocol, which will not exceed 200 no matter how the congestion window grows. Such limitation largely reduces the throughput rate of QUIC protocol in the network environment with high bandwidth and long round trip delay.

公开(公告)号：US20170337686A1

公开(公告)日：2017-11-23

申请号：US15598656

申请日：2017-05-18

Applicant: SICHUAN UNIVERSITY

Inventor： Junfeng WANG ,  Lin GAO ,  Fan LI ,  Yulin JI ,  Zongan LIANG

IPC: G06T7/00 ,  G06T7/11

CPC classification number: G06T7/0012 ,  G06T5/10 ,  G06T5/20 ,  G06T7/11 ,  G06T7/136 ,  G06T7/168 ,  G06T7/262 ,  G06T2207/10116 ,  G06T2207/20016 ,  G06T2207/30008 ,  G06T2207/30061

Abstract: A X-ray chest image rib suppression method based on Poisson model. It conducts contourlet transformation on the image and utilizes transformation coefficient correlation between different scales to conduct texture enhancement on the image; it designs strip-type detection filter in accordance with the Hessian matrix eigenvalue to the image and detects the area where the rib locates in; it combines enhanced texture and rib area information, establishes and solves rib suppression Poisson model, realizing the rib suppression in the image. Anisortropy and contourlet transformation multi-direction feature is utilized, scale and coefficients direction information are combined and distinction degree between texture and noise improves, enhancing texture while restraining noise; it realizes ribs suppression through solving the Poisson model, which does not need to conduct accurate segmentation on the rib, prevents unnatural transition problem of edges resulted from explicit ribs suppression and effectively suppress the ribs, improving observation effect of X-ray chest image.

公开(公告)号：US20190190542A1

公开(公告)日：2019-06-20

申请号：US16219868

申请日：2018-12-13

Applicant: SICHUAN UNIVERSITY

Inventor： Junfeng WANG ,  Qin LUO ,  Ying LIU ,  Hao NIU

IPC: H03M13/37 ,  H04L12/26 ,  H04L1/00 ,  H04W84/18 ,  H04L12/801

CPC classification number: H03M13/3761 ,  H04L1/0057 ,  H04L43/0841 ,  H04L43/0864 ,  H04L47/34 ,  H04W84/18

Abstract: This invention discloses a kind of reliable data transmission method based on reliable UDP and fountain code in aeronautical ad hoc networks. This method employs Raptor codes to recover from data loss and avoid retransmission in high loss and high delay channels. A light-weight feedback scheme is designed to provide reliability. An improved SNACK scheme is utilized to deal with the bandwidth asymmetric problem and a novel sliding window mechanism is proposed to avoid TCP's retransmission ambiguity problem and allow for more precise roundtrip-time calculation. A congestion control scheme is deployed to distinguish losses caused by congestion from losses caused by link handoffs to efficiently utilize the available bandwidth. This invention is fully implemented directly on top of UDP in Linux without requiring changes to the underlying network stack implementations. In harsh aeronautical ad hoc network environment, it is suitable for reliable and efficient data transfer service.

公开(公告)号：US20180144132A1

公开(公告)日：2018-05-24

申请号：US15630450

申请日：2017-06-22

Applicant: SICHUAN UNIVERSITY

Inventor： Junfeng WANG ,  Xiaosong ZHANG ,  Yao DU ,  Jie LIANG ,  Yong MA

IPC: G06F21/56 ,  G06Q10/04

CPC classification number: G06F21/563 ,  G06F11/3676 ,  G06F17/16 ,  G06F17/2785 ,  G06F21/561 ,  G06N99/005 ,  G06Q10/047

Abstract: This invention comes up with a kind of Android malicious code detection method on the base of community structure analysis. During the reverse analysis process of target program, firstly, it obtains critical static feature information automatically, such as permission, function, class, system API, etc,; secondly, it uses the call relation between functions to create function call graph, and undertakes pretreatment on function call graph; make cycle division and analysis for the weighted function call graph so as to get the correction division of community structure; finally, it extract features from community structures for machine learning and get the final maliciousness determination result. This invention method is able to undertake program internal structure analysis and malicious code detection rapidly when facing a large number of Android application program samples generated by “repackaging” technology.

公开(公告)号：US20180130202A1

公开(公告)日：2018-05-10

申请号：US15598628

申请日：2017-05-18

Applicant: SICHUAN UNIVERSITY

Inventor： Junfeng WANG ,  Peng TANG ,  Fan LI ,  Yihua Du ,  Yulin JI ,  Zongan LIANG

IPC: G06T7/00 ,  G06T7/13 ,  A61B6/00

CPC classification number: G06T7/0012 ,  A61B6/50 ,  A61B6/5211 ,  A61B6/5217 ,  G06T7/13 ,  G06T2207/10116 ,  G06T2207/20081 ,  G06T2207/20084 ,  G06T2207/30061

Abstract: A DR radiography lung contour extraction method based on fully convolutional network, which includes the steps: Establish the fully convolutional network structure of lung contour segmentation; Conduct off-line training on the weighting parameters of the fully convolutional network; Read DR image and weighting parameters of the fully convolutional network; Input DR image into fully convolutional network and output segmentation results of image through network terminal with network layer-by-layer feedforward. Establish lung contour in accordance with segmentation results.

公开(公告)号：US20180121652A1

公开(公告)日：2018-05-03

申请号：US15645548

申请日：2017-07-10

Applicant: SICHUAN UNIVERSITY

Inventor： Junfeng WANG ,  Jie LIANG ,  Xiaosong ZHANG ,  Dong LIU ,  Yong MA

IPC: G06F21/56

CPC classification number: G06F21/566 ,  G06F21/56 ,  G06F21/567 ,  G06F2221/033

Abstract: A kind of malicious software clustering method expressed based on TLSH feature, which belongs to analysis and test area of malicious software. The Cuckoo Sandbox is used to analyze malicious software to acquire three kinds of character string features, which are static feature of software, resource assess record during operation and API; then character strings are disassembled, filtered and sorted and TLSH algorithm is used to compress them into three groups of feature values with size of 70 characters; Finally, OPTICS algorithm is utilized to realize automatic classification on malicious software family. Invention adopts unsupervised learning methods, which does not need manual tab for training in advance. Features which are extracted are compressed and expressed by using TLSH. Under situation that feature is not lost, data dimension is largely lowered and clustering speed is improved.

公开(公告)号：US20180082064A1

公开(公告)日：2018-03-22

申请号：US15645767

申请日：2017-07-10

Applicant: SICHUAN UNIVERSITY ,  BEIJING TONGTECH CO., LTD.

Inventor： Junfeng WANG ,  Baoxin XU ,  Dong LIU ,  Fan LI ,  Xiaosong ZHANG

IPC: G06F21/56 ,  G06N99/00

CPC classification number: G06F21/566 ,  G06F21/563 ,  G06F2221/033 ,  G06N5/003 ,  G06N7/005 ,  G06N20/00

Abstract: A method of detecting malware in Linux platform through the following steps: use objdump-D command to disassemble ELF format benign software and malware samples to generate assembly files; traverse the generated assembly files one by one, read the ELF files' code segment and meanwhile identify whether the code segment contains main( ) function; analyze the code segment read. Divide assembly code into different basic blocks. Each basic block is marked by its lowest address. Add control flow graph's vertex to the adjacency linked list; establish the relation between basic blocks, add control flow graph's edges to the adjacency linked list and generate a basic control flow graph; extract control flow graph's features and write them into ARFF files; take ARFF files as the data set of a machine learning tool named weka to carry out data mining and construct classifier; classify the ELF samples to be tested by using the classifier.