82 results about "Deep integration" patented technology

The invention discloses a deep learning-based multiview face three-dimensional model reconstruction method, and belongs to the field of computer vision. The method comprises the following steps of: generating a multi-illumination multiview virtual face image; generating a depth map of a face front view; training a plurality of independent and parallel convolutional neural networks; training a neural network in which weights of various views are distributed; and restoring depth maps output by the networks into a face three-dimensional grid model and carrying out peak coloring. According to themethod, multiview images are independently trained to restore depth maps, and each view weight distribution map is trained to carry out deep integration, so that the face three-dimensional model reconstruction precision is improved under the premise of ensuring the efficiency.

The invention discloses a SINS / CNS deep integrated navigation system and a realization method thereof. Wherein the navigation system comprises a strapdown inertial navigation system (SINS), a celestial navigation system (CNS), an integrated navigation filter, a inertial navigation posture measurement information structure unit; the realization unit comprises the following steps: 1. a large viewing field star sensor assists the strapdown inertial navigation system to obtain a high-precision mathematic horizontal reference; 2. CNS positioning is carried out based on the mathematic horizontal reference; 3. SINS / CNS deep integrated system model and measurement mode are established; 4. integrated navigation system information is fused; 5. the SINS and the CNS assists each other to realize high-precision positioning. In the invention, the star sensor high-precision posture information is employed to assist SINS to obtain high-precision SINS strapdown matrix which serves as the mathematic horizontal reference for CNS positioning, and on the basis, positions and postures of the CNS are employed to comprehensively calibrate the SINS, thus realizing SINS / CNS deep integration and finally achieving high-precision positioning and navigation.

The invention discloses a satellite signal vector tracking method based on SINS / GPS deep integration data fusion. In the method, compensation is carried out on inertia element errors through channel filter information fusion, SINS / GPS deep integration navigation data fusion and an SINS system based on an SINS / GPS deep integration navigation system; and position and speed parameters of a current computation period are corrected to obtain accurate navigation parameters and generate local inference signals as well as keep the tracking for all visual GPS satellite signals; and a loop formed by channel filters and a main filter is utilized to complete the vector tracking and integration navigation information processing to the GPS satellite signals. The invention has excellent noise inhibitioncapacity and dynamic tracking performance, can ensure the navigation accuracy and the reliability of an integration system during the temporary interruption of the GPS satellite signals, and can maintain better pseudo code phase and carrier frequency tracking property in an environment with low carrier-to-noise ratio.

The invention relates to a speedometer-assisted vehicle-mounted dynamic alignment method. The method is technically characterized by comprising steps as follows: in an inertial navigation system, inertial navigation calculation is performed, and attitude heading, speed and position information is output and updated; mileage decomposition is performed by use of an attitude angle of vehicle displacement increment output and inertial navigation calculation of a speedometer, and speedometer navigation position calculation is completed; an error model of the inertial navigation system and an errormodel of navigation position calculation are taken as equations of propagation of an error state, an attitude angle error of the inertial navigation system is estimated and calibrated by use of a Kalman filter. Deep integration is performed through the Kalman filter by use of displacement increment information provided by the vehicle-mounted speedometer and gyroscope and accelerator data providedby an inertial measurement assembly, in the running process of a vehicle, a combined navigation technology of accurate alignment of attitude heading and correction function of position errors is completed, the starting process of the vehicle-mounted navigation system can be simplified, the reaction time is shortened and the maneuvering characteristics of a carrying vehicle are improved.

The invention discloses a smart city internet of things system based on a cloud platform. The smart city internet of things system comprises a cloud platform data center, an external equipment connecting and converting port, a face recognition system, a vehicle track management system and a chemical park security management and pre-warning system. The external equipment connecting and converting port is used for connecting external equipment to realize transmission and intercommunication of resources; the face recognition system realizes locating and real-time monitoring on suspicious people; the vehicle track management system monitors vehicles in real time and is connected with a police geographic information system through the cloud platform data center so as to find dispatching or regulation-violating vehicles; and the chemical park security management and pre-warning system inputs types, features, leakage or discharge positions of common toxic pollutant into the cloud platform data center which locates warning position and probable diffusion range according to signals sent by a sensor and timely adopts emergency measures. The smart city internet of things system carries out deep integration based on cloud computation technology and integrates subsystems into a uniform platform so as to avoid reconstruction; and the smart city internet of things system provided by the invention has significance for security, effective management and environmental protection of a city.

The invention discloses a geographic grid-based government affair information resource integration method. The method comprises the steps of performing multilevel geographic grid division; performing spatialization processing of government affair information resources; performing time-space coherence processing of the government affair information resources; and performing information integration of multi-scale geographic grids. By fully utilizing a geographic grid division technology, a grid coding method, a text word segmentation and extraction method and a spatial statistical analysis technology, the fusion of government affair information and space information is realized; deep integration and application of the government affair information resources are performed; and grid analysis of single type of data is upgraded to grid internal integration analysis of government affair information resources of multiple types and multiple scales, so that comprehensive government affair data space distribution, comparison and associative analysis can be carried out in a specified region. The information extraction and resource integration can be quickly realized in regional planning and emergency of governments.

The present invention provides a system and method for aircraft to determine own position and navigate using a navigation heartbeat signal broadcast on a DME uplink and / or a Mode-S uplink frequency. The present invention enables deep integration between the existing navigation systems (DME interrogation-reply ranges and GPS / WAAS raw TDOA or pseudo range measurements) and the DME heartbeat TDOAs or Mode-S heartbeat TDOAs to provide a highly accurate navigation positioning capability and provide necessary backup capability in lieu of GPS to maintain the necessary RNP / RNAV capability and avoid degrading aircraft operational safety.

The invention relates to a complete network middleware operation support platform which can be used for integrating a plurality of conventional mainstream middleware technologies, and a method for integrating a heterogeneous middleware member on the platform. The platform disclosed by the invention comprises a microkernel integration frame, a communication service layer, a public service layer, a member container layer and a service engine layer, wherein the microkernel integration frame provides upper-level service integration interface specifications of a life cycle management and command interface, an interoperation interface and a deployer interface needed for deploying specific components, for the four categories of service layers. According to the invention, deep integration of mainstream middleware technologies, such as JEE (Java Enterprise Edition), Web service, CORBA (Common Object Request Broker Architecture), COM / DCOM (Component Object Model / Distributed Component Object Model), and the like, is realized. The complete network middleware operation support platform and the integration method of the heterogeneous middleware can be widely applied to integration of different application service technology systems.

The invention relates to a rail subgrade settlement measuring method based on combination of camera shooting, a satellite and inertia measurement. A vehicle platform is provided with a camera, a satellite positioning assembly, an inertia measurement assembly, a time synchronous system, a rail vehicle relative measurement system, a data processing special-purpose computer and the like. Satellite positioning differential base stations and camera shooting survey mark base stations are distributed along a rail to constitute a combination measurement system of camera shooting, the satellite and inertia measurement or a combination measurement system of camera shooting and inertia measurement or a combination measurement system of the satellite and inertia measurement, high-precision vehicle-mounted dynamic position, altitude and speed measurement references are built through deep integration of camera shooting, the satellite and inertia measurement, high-precision measurement of rail and subgrade measurement parameters is maintained for a long time according to the rail vehicle relative measurement system, and by comparison with rail design standard data, automatic, high-precision, long-time and continuous measurement of rail subgrade settlement and rail geometric parameters is achieved.

The invention relates to a global smart manufacturing sales service system integrating information technology, Digital and intelligent global purchasing, Payment, Intelligent Manufacturing, sales andlogistics distribution as one, enabling Global Power, Small and medium-sized enterprises, making them available for global procurement, Global sales, Intelligent manufacturing capabilities, and improving global logistics and distribution, efficiency of global payments. At the same time, in order to ensure the independence of enterprise operation information, based on confidentiality and security,global enterprises can share the experience of intelligent manufacturing, continuously improve product quality, Reduce operating costs, deep integration of end customers and manufacturers can also beachieved, on the basis of big data analysis, the global enterprises can adjust their business decisions in real time, and can accurately push the product information to the customers to improve customer satisfaction and shopping experience, which greatly improves the efficiency of the supply chain, reduces the purchase cost of end-user products, and reduces the operating costs of enterprises, andhas great social benefits.

The invention discloses an inertia-assisting GPS deep-integration semi-physical simulation system which comprises a GPS receiver, a satellite navigation signal simulator composed of a control mainframe and a radio frequency signal simulator, a ballistic trajectory generator, a strapdown inertia navigation system information simulator composed of an IMU simulator and an inertia navigation computer, a simulation analysis computer and a signal forwarder. The strapdown inertia navigation system information simulator processes the simulation information output by the ballistic trajectory generator to obtain inertia navigation data and transmits the inertia navigation data to the GPS receiver. The ballistic trajectory generator transmits guided missile flight path data to the radio frequency signal simulator. The control mainframe controls the radio frequency signal simulator to generate a satellite navigation simulation signal and sends out the satellite navigation simulation signal through the signal forwarder. The GPS receiver processes the inertia navigation data and the satellite navigation information comprehensively and outputs the final navigation information to the simulation analysis computer. By means of the inertia-assisting GPS deep-integration semi-physical simulation system, the influence of inertia-assisting on the acquisition tracking performance of the GPS receiver is analyzed, and a great significance is provided for research of the deep-integration navigation technology.

The invention provides a fixed-ambiguity PPP (Precise Point Positioning) / SINS (Starpdown Inertial Navigation System) tight combination positioning and attitude determining method, which is characterized in that a PPP / SINS tight combination mode is adopted, information deep integration is performed at an original observed value level, PPP floating point ambiguity calculation is assisted by using the high-precision location obtained by using inertial navigation in a recursive manner within a short time, the PPP ambiguity is fixed after the high-precision floating point ambiguity is obtained, the wide-lane ambiguity and the narrow-lane ambiguity are sequentially fixed, the reset of the state parameters are updated again by using the successfully fixed narrow-lane ambiguity, the ambiguity is kept to be fixed by adopting a transfer mode, and continuous high-precision positioning and attitude determining are realized. The method provided by the invention can significantly improve the accuracy and the reliability of PPP / SINS combined positioning and attitude determining, re-convergence and re-fixing of the PPP ambiguity can be accelerated particularly under a condition of GNSS signal interruption, and the availability of a PPP / SINS combination technology in a complex environment is enhanced.

The invention provides a network security authentication method. The network security authentication method comprises the following steps: S1, configuring a KDC server based on Kerberos authentication service from a web client by an administrator; S2, correlating resources with a user Id, a role, a Token Id and a Token state to construct correlation information by the KDC server; S3, acquiring a token value and token information from the web client by a user, saving the token information in a persistent node, and binding a token to the resources; S4, according to the token information and the correlation information in the step S2, determining the right for the user to access the resources. The user can configure the KDC server through a web browser, so that one-key data deployment security service is achieved, and effective monitoring on the running state of the KDC server is achieved, deep integration with large data and cloud computation is achieved, and thus a protecting mechanism with a higher security level is provided for the resources of the user.

The invention discloses a knowledge graph construction method based on an improved BERT model, and the method comprises the steps: obtaining multi-source heterogeneous data, and storing the obtained multi-source heterogeneous data to a big data platform; carrying out knowledge modeling on the acquired multi-source heterogeneous data based on the ontology; acquiring a knowledge on the basis of a BERT-BiGRU-CRF model; performing entity fusion and attribute fusion on the obtained heterogeneous data sources based on knowledge fusion of word vectors; and performing knowledge storage by adopting a graphic database. The enterprise knowledge graph is adopted to break through a traditional calculation mode, internal and external information and data in the field of enterprise operation management are deeply integrated, early warning potential risks can be more effectively mined in the later period, enterprise operation management efficiency is improved, and improvement of accuracy of subsequententerprise operation management risk intelligent diagnosis is facilitated.

The invention relates to a high-precision Beidou tri-band SINS deep integration navigation system, wherein the Beidou receiver navigation error eliminating method and the ambiguity solving method are considered, and a Beidou navigation satellite system and an SINS help each other for combined navigation. Through the TCAR method, the whole cycle ambiguity is accurately and efficiently solved, and therefore the high-precision carrier phase measurement is conducted; through tri-band data signals, ionospheric refraction errors can be solved according to the equation set composed of ionospheric refraction error calculation formulas, and then the positioning precision is improved. The system focuses on solving the problem that the deep integration navigation system relies on the accuracy of the SINS, the navigation precision of a receiver is improved through Beidou tri-band data fusion, the requirements for the posture, speed and position precision of the SINS is lowered, the anti-interference capacity and the comprehensive accuracy of the Beidou tri-band receiver and SINS combined navigation system are improved, and the performance of armament and important civil carrying tools can be improved.

The invention provides a satellite-rocket integrated multifunctional structure aircraft for ground simulation and relates to the technical field of aerospace structure design. The satellite-rocket integrated multifunctional structure aircraft mainly comprises a propulsion subsystem, a common multifunctional solar panel, a common multifunctional service module and an effective load, wherein the effective load is arranged at the front end; the common multifunctional service module is located in the middle; the propulsion subsystem is located at the bottom of the common multifunctional service module; and the rear end of the system is also connected with an adapter of a vehicle. Through designing a series of novel concept multifunctional structures, such as the common multifunctional solar panel, a multifunctional structure cable, a multifunctional satellite-rocket adapter, a multifunctional structure battery and a multifunctional structure computer, deep integration of the structures is achieved, and the structure quality and redundancy volume of a system are effectively reduced, so that the payload ratio of a satellite is greatly improved, and the performance of the satellite-rocket integrated aircraft is greatly strengthened.

An alarm message center system for regional power generation company is composed of a data source layer, a data service middleware layer, an alarm rule configuration layer, an alarm task calculation engine layer, a message bus and a message distribution center layer, an alarm display and analysis layer. Visual component configuration of alarm rules, distributed computing engine of alarm task, multi-terminal message distribution center, multi-user scenario alarm function analysis and display are realized. Through the alarm message center system, the production management personnel of the regional power generation company and the relevant responsible persons of the power plant can flexibly formulate the alarm business rules, obtain the alarm message notification at the first time and timelygrasp the operation status of the generator set, so as to maximize the reaction speed of the accident and reduce the possibility of the accident. At the same time, it can analyze the alarm equipment afterwards to provide decision-making basis for equipment early warning, form closed-loop management, and reflect the new thinking of deep integration of informationization and industrialization.

The invention discloses a method and system for deep integration and real-time compensation of micro-inertial measurement information. The method comprises the following steps: respectively acquiring angular velocity information, acceleration information and magnetic field intensity information of a study object; respectively carrying out real-time compensation on the angular velocity information, the acceleration information and the magnetic field intensity information by virtue of a deep learning method, so as to respectively obtain compensated angular velocity information, acceleration information and magnetic field intensity information; determining a first Euler angle according to the compensated angular velocity information, and determining a second Euler angle according to the compensated acceleration information and the compensated magnetic field intensity information; and determining an integration gain coefficient of a self-adaption gain method according to a mean square error of the first Euler angle and a mean square error of the second Euler angle, and integrating the first Euler angle with the second Euler angle according to the integration gain coefficient, so as to obtain the integrated Euler angle. According to the method and the system, the micro-inertial measurement precision can be increased, the stability and the reliability are high, the aging ratio is high, and the response speed is high.

The invention provides a head-up display system and a control method thereof, and relates to the field of display technologies. According to the invention, the conversion between light with differentpolarization states is performed through a polarization conversion element, images of a first projection plane and a second projection plane are simultaneously displayed by using a property of playingan imaging effect on light with an S polarization state and light with a P polarization state of a birefringent lens assembly and a scattering effect of a scattering assembly, and the light field display is realized through deep integration of the images of the two projection planes, thereby being capable of providing eye focusing adjustment within a certain depth range for a driver, displaying anatural 3D image, enabling the displayed image to be more authentic, and avoiding eye fatigue caused by long-term viewing of the driver.

The invention belongs to the technical field of computer network and communication and discloses a medical health care management information service system. The system includes a software system including a service processing subsystem, a user service subsystem, a service management subsystem, a system management subsystem and a user WEB service subsystem, a user health medical treatment terminal, a presentation layer subsystem, a management layer subsystem, a communication layer subsystem, a domain control subsystem and an agent layer subsystem. According to the medical health care management information service system, a computer software development technology, an information network technology and a related hardware and software support technology are applied, the deep integration ofdifferent communities and medical institutions is achieved in a centralized operating environment to meet the modern management needs of the medical institutions for community health information services, medical monitoring is implemented, and the medical service industry is standardized; and the problems that grass-roots doctor have a diagnosis and treatment level, the medical and health industryinformatization cost is high, the public medical resource utilization rate is low and human resources in the medical and health industry are lacked are solved.

The invention relates to a navigation method based on polarization-geomagnetic vector tight integration. The navigation method based on the polarization-geomagnetic vector tight integration comprisesthe following steps: firstly, intelligently extracting solar vector information under a carrier coordinate system by using an imitating compound eye polarization compass; secondly, acquiring the solarvector information under a geographic coordinate system in combination with solar astronomical ephemeris, and establishing an information conversion relation of a solar vector under a body coordinatesystem and the geographic coordinate system; extracting a geomagnetic vector under the body coordinate system and the geographic coordinate system, and establishing a conversion relation of the geomagnetic vector under the body coordinate system and the geographic coordinate system; then carrying out information integration on the geomagnetic vector and the solar vector under the body coordinatesystem; and finally, establishing a combined navigation system model based on polarization / geomagnetic vector integration. Compared with an existing method, the method has the advantages that deep integration of geomagnetic information and polarization information is realized, the system precision and stability can be improved, and the navigation method can be used for the fields of attitude determination and positioning and the like of carriers such as a ship, an unmanned aerial vehicle, a guided missile and an unmanned vehicle.

The invention provides an energy information processing method and equipment and an energy internet system. The energy equipment comprises an interface module, an energy state identification module and a power consumption adjusting module; the interface module is configured to receive energy information reported by other energy equipment in the system, and the energy equipment in the system can automatically report energy information when accessing the energy internet system or quitting the energy internet system or in the process that energy data changes; the energy state identification module is configured to judge whether or not the system is in an energy shortage state according to the current energy supply condition of the system after the receiving module receives the energy information; the power consumption adjusting module is configured to reduce power consumption of the energy equipment on the condition that the system is in the energy shortage state. Accordingly, by conducting deep integration and mutual assistance on energy and information, all parts in the system can respond to system energy change in real time, and a distributed local energy internet system is built to form brand new energy ecology.

The invention discloses a multi-information deep integration navigation micro-system and a navigation method. The micro-system comprises a three-axis MEMS (Micro Electro Mechanical System) gyroscope chipset, a three-axis acceleration chipset, a satellite navigation chip, a magnetic field meter, a height meter, a signal-driven demodulation amplifying circuit, a multi-channel ADC (Analog to Digital Converter) chip, a navigation solution circuit and a power supply management circuit. The navigation method comprises the following steps: (1) initializing the micro system; (2) acquiring a motion state signal of a carrier; (3) receiving a satellite navigation radio frequency signal; (4) performing satellite capturing, satellite loop tracking and satellite navigation solution; (5) judging an initial alignment state; (8) if initial alignment is finished, performing a strap-down inertial navigation calculation; (9) judging whether satellite navigation information is available or not; (10) acquiring an optimal navigation parameter, and meanwhile feeding back the optimal navigation parameter. The navigation system has more functions and higher navigation accuracy, is miniaturized and is low in power consumption and cost, so that the demand of modern navigation is met.

The invention discloses a deep integration target tracking method based on a time and space network. The method comprises: using a VGG-16 network to extract characteristics of third, fourth, and fifthlayers, expressing the traditional correlation filtering algorithm as a layer of convolutional neural network to obtain a correlation filtering network; a time network and a spatial network are constructed on the basis of a related filtering network, so that time information and space information of the target are further captured, improving the accuracy of an algorithm. Meanwhile, all weak trackers are fused into strong trackers through an adaptive weight integrated learning algorithm to achieve target tracking, integrated learning enables the method to have good robustness, target trackingunder complex scenes can be coped with, finally, an updating strategy combining short-time updating and long-time updating is provided, and the stability of the model is ensured.

The invention discloses a Beidou inertial navigation deep integration navigation microsystem. The Beidou inertial navigation deep integration navigation microsystem comprises a three-axis micromechanical gyroscope, a three-axis micromechanical accelerometer, an odometer encoder, an altimeter, a magnetic field meter, a Beidou / GPS dual-mode chip, a signal conversion and interface circuit, a digital signal processor and a secondary power chip, wherein the digital signal processor is used for receiving digital signals converted by the signal conversion and interface circuit and calculating digital signals corresponding to the angular speed and the accelerated speed so as to obtain the speed and the displacement of a carrier, and the digital signal processor is further used for carrying out the Kalman filtering algorithm on digital signals corresponding to the speed and the displacement of the carrier, travel information, height information, magnetic field information, Beidou signals and GPS signals so as to obtain navigation parameters. According to the Beidou inertial navigation deep integration navigation microsystem, a Beidou, the gyroscope, the accelerometer, the magnetic field meter, the altimeter and functional chips of signal conversion, digital signal processing and the like are organically integrated, full-parameter navigation and seamless navigation can be achieved, and interference resistance and application reliability of the whole navigation system are improved.

The invention discloses a method and a device for converting a planar video into a stereoscopic video. The method comprises the following steps executed on each frame image of the planar frame: S1, acquiring a depth map D of a current frame image: acquiring a first depth map D1 through motion estimation based on block matching; constructing a second depth map D2 based on a geometric perspective relation through an edge detection algorithm and a Hough transform algorithm; estimating a third depth map D3 through a method based on color information; and executing deep integration on D1, D2 and D3 to obtain the depth map D; S2, generating a multi-viewpoint stereoscopic view through a reference map and the depth map D based on a DIBR (Depth Image Based Rendering) algorithm; and S3, selecting at least a part of left-eye and right-eye views from the multi-viewpoint stereoscopic view according to the stereoscopic video format requirement of a user for executing stereoscopic rendering to generate a color stereoscopic video of a corresponding format. The stereoscopic video generated through the method has a good stereoscopic effect, and stereoscopic videos of different formats can be generated according to user demands.

The present invention provides a system and method for aircraft to determine own position and navigate using a navigation heartbeat signal broadcast on a DME uplink and / or a Mode-S uplink frequency. The present invention enables deep integration between the existing navigation systems (DME interrogation-reply ranges and GPS / WAAS raw TDOA or pseudo range measurements) and the DME heartbeat TDOAs or Mode-S heartbeat TDOAs to provide a highly accurate navigation positioning capability and provide necessary backup capability in lieu of GPS to maintain the necessary RNP / RNAV capability and avoid degrading aircraft operational safety.

The invention provides a method and a system for bonding fixed-line extension telephones and mobile terminals. The method comprises the following steps of: establishing SIP UA (Session Initiation Protocol User Agent) connections between soft switch equipment and an IP (Internet Protocol) private exchange according to the requirement of a voice concurrency amount and assigning a soft switch number for each SIP UA connection as an outside line number of a fixed-line telephone; setting fixed-line extension telephone numbers under the outside line number of the fixed-line telephone, wherein each fixed-line extension telephone number corresponds to a fixed-line extension telephone; configuring mobile terminals to be bonded to the fixed-line extension telephones, storing a corresponding relation between the mobile terminal numbers and the fixed-line extension telephone numbers in the IP private exchange and storing a corresponding relation between the mobile terminal numbers and the outside line number of the fixed-line telephone in an application server; sending registration information to the soft switch equipment by the IP private exchange in an SIP UA mode; and configuring business subscription information for the bonded the mobile terminal numbers in a network to which the mobile terminals belong. The invention realizes the access of an IP PBX (Private Branch Exchange) into a telecommunication network and the deep integration application of the telecommunication network and an enterprise network.

The invention discloses a reliable and controllable vehicle electronic identity information acquisition device comprising an electronic identifier reading module, a satellite positioning module and a wireless communication module which are bound to one another, wherein the output end of the vehicle electronic identifier reading module and the output end of the satellite positioning module are connected with the wireless communication module. The device is used for acquiring vehicle electronic identity information, traveling track and vehicle running state information in real time and uploading the information automatically, so as to realize deep integration of the vehicle electronic identifier and the satellite positioning module, effectively prevent the problem of identity management chaos caused by satellite positioning terminal disassembly and replacement without permission and improve effective regulation of operating vehicles by the public security organ and the transportation department and the timeliness of information acquisition. The device can also be applied to vehicle-to-vehicle communication (v2v) identity authentication management of intelligent networked vehicles.

The invention discloses a pulmonary nodule feature extraction method based on a convolutional neural network and principle component analysis, which comprises the steps of firstly automatically learning pulmonary nodule features by using the convolutional neural network, performing feature extraction through convolution, and performing feature mapping through downsampling; and secondly, performing dimension reduction on the output of each feature mapping layer in a convolutional neural networks model for feature extraction (FeCNN) by using PCA (Principal Component Analysis) in feature extraction, and integrating with mapping of an output layer to acquire final multi-layer deep integration features. The pulmonary nodule feature extraction method not only can effectively recognize medical symptoms of a pulmonary nodule, but also avoids the complex feature extraction and feature reconstruction process in a traditional method, and plays a role of assisting diagnosis from an objective aspect.