91results about How to "Add scene" patented technology

The invention discloses a human body behavior identification method and system based on multi-target detection 3D CNN, and the method comprises the steps: 1) carrying out the preprocessing of a video,and enabling the video flow to be converted into an image frame; 2) carrying out calibration cutting on a target object in the video by adopting a relatively mature SSD detection technology at present; 3) establishing a feature extraction network structure of the image frame data and the calibration clipping data; 4) establishing a feature fusion model, and fusing the two features extracted in the step 3); 5) carrying out classification by using a Softmax regression model classifier; and 6) carrying out fine tuning on the trained model according to an actual application scene or a public dataset. According to the method, a condition of information loss caused by convolution of the current deep neural network model in the time dimension is made up, the expression of features in the time dimension is enhanced, the recognition efficiency of the model is integrally improved, and the model can better understand the behavior action of a human body.

The invention discloses a somatosensory movie playing system which comprises a virtual reality terminal, a virtual reality platform and a data support server. The virtual reality terminal comprises a visual display module and a man-machine operation interface, and is used for displaying movie category guidance and browsing and auxiliary information of movie contents and providing a man-machine interface. The virtual reality platform is used for providing hardware and software platforms needed by system operation, virtual reality application programs are installed in the virtual reality platform, and the virtual reality platform receives operation requests sent by the virtual reality terminal, reads data from the data support server and displays the data on the virtual reality terminal. The data support server is used for providing data information needed by a movie system. The somatosensory movie playing system has the advantages that original reading modes of electronic movies are broken through, and by an immersive reading mode, reading quality is further improved, and better user experience is provided.

An automated parking device detects a peripheral environment around a periphery of a host vehicle including parking spaces having obstructions therein, and determines, based on the peripheral environment around the periphery of the host vehicle, a degree of obstruction when parking is attempted in the parking spaces, the automated parking device including selection units which enable the selection of parking in desired parking spaces, and a vehicle control unit which, when a user has selected to park in the parking space having a small inhibitory factor, executes an automated parking operation by causing the user to recognize that parking in the parking space is being performed.

The invention relates to the technical field of automobile safety control, in particular to an emergency lane keeping assisting method for a VRU. After it is judged that an own vehicle has the tendency of deviating from an own lane, VRU target information and lane line information around the own vehicle are collected, whether the current situation of the own vehicle meets the starting condition ofemergency own vehicle deviation rectification or not is judged according to the VRU target information and the lane line information around the own vehicle, if the current situation of the own vehicle meets the starting condition, whether the own vehicle possibly collides with the target VRU or not according to the own vehicle deviation track, if collision possibly occurs, an own vehicle deviation rectifying track for avoiding collision between the own vehicle and the target VRU is obtained, whether the own vehicle deviation rectifying track meets set requirements or not is judged, and if theown vehicle deviation rectifying track meets the set requirements, deviation is rectified according to the own vehicle deviation rectifying track; and the target VRU is the VRU located in a set areaof the own vehicle. According to the method, the vehicle is prevented from colliding with VRUs at the two sides of the own lane, and the driving safety of the vehicle is improved to a great extent.

The invention relates to the technical field of intelligent electronic equipment, in particular to AR glasses. The AR glasses comprise a display module; an adjusting mechanism; a spectacle frame, wherein the display module and the adjusting mechanism are arranged on the spectacle frame; the adjusting mechanism is arranged to be capable of adjusting the image of the display module to be within a set visual field range. The AR glasses provided by the invention can meet the imaging effect of people with different head circumferences and pupil distances when the people wear the AR glasses.

The invention discloses an object grabbing method of a robot. The method includes the steps that a first position of the robot and a second position of an object are obtained, wherein the first position and the second position come from Kinect infrared equipment; a first motion trail from the robot to the object is obtained according to the first position and the second position; and a target position is obtained according to the first motion trail so that the robot can move to the target position to grab the object. By means of the object grabbing method, the object positioning accuracy is achieved, and therefore the beneficial effects of grabbing the object more accurately and enlarging the using scene and scope of the robot are achieved, and the technical problems that in the prior art, a method for positioning by using self vision of the robot is low in positioning accuracy and not suitable for the environment with dark light are solved.

The invention provides a human body tone detection method and device. The human body tone detection method includes: automatically opening a tone detection sensor when a microphone is opened by voice input, wherein the tone detection sensor is integrated in the microphone; acquiring input voice features through the microphone, and acquiring a tone that a user outputs through the tone detection sensor; analyzing a current tone condition based on the voice features and the tone outputted by the user; and prompting the analyzed current tone condition. According to the method and the device, the microphone and the tone detection device are integrated; when the user speaks through the microphone, tone detection can be achieved; application is more natural, and tone detection will not be forgot; the device can timely notify the user of the current tone condition; and the human body tone detection device is integrated with the microphone and can be directly used in all the voice input mobile devices, and the user scenes and the use range can be expanded.

The invention provides a high-strength stainless steel hollow composite plate for building construction or template. The composite plate comprises a stainless steel surface plate, a stainless steel bottom plate, a stainless steel honeycomb plate and stainless steel side plates, wherein the stainless steel honeycomb plate is arranged between the stainless steel surface plate and the stainless steel bottom plate, and the stainless steel honeycomb plate is obtained by adopting mold pressing integrated molding; and the stainless steel side plates are arranged between the stainless steel surface plate and the stainless steel bottom plate for edge sealing.

The invention discloses a construction method of an intelligent assembly system based on digital twinning, and belongs to the technical field of intelligent manufacturing, and the method comprises the steps: carrying out the real-time collection of the dynamic data of an assembly process of a physical workshop through a data collection system according to the dynamic data elements of the technical physical workshop of a complex equipment product assembly production site; pre-processing and standardizing, and storing into a database; a digital twin workshop model is used as a system client, and through interaction with a database, unified data service drives an assembly part three-dimensional virtual model and a product three-dimensional model, so that bidirectional dynamic mapping between a digital twin workshop and an assembly production physical workshop is realized; based on three-dimensional visualization system integration, construction of an intelligent workshop digital twinning system is completed, so that the functions of real-time state display, state evolution, data input, data storage and the like in the digital aircraft assembly process are achieved, and the problems that existing digital workshop data are monotonous in presentation and poor in interaction and immersion are solved.

The invention provides an unmanned aerial vehicle take-off method and device, a storage medium, an autopilot and an unmanned aerial vehicle, and relates to the technical field of flight control. A received take-off instruction is responded to adjust output power of the unmanned aerial vehicle so as to enable the unmanned aerial vehicle to reach set flight parameters when executing the take-off instruction, and a locking mechanism is controlled to unlock when the unmanned aerial vehicle is determined to reach a take-off condition of a take-off platform so that the unmanned aerial vehicle takesoff from the take-off platform according to a set climbing speed and the set flight parameters. Compared with the prior art, by using the method, the device, the medium, the autopilot and the unmannedaerial vehicle of the invention, when the unmanned aerial vehicle takes off, take-off operation can be executed without being limited by a state of the take-off platform, scene condition limitation when the unmanned aerial vehicle executes the take-off operation is relieved, and a take-off execution scene of the unmanned aerial vehicle is expanded.

The invention relates to a flame vision distance measuring device and method. The flame vision distance measuring device comprises a camera used for photographing a flame and a controller used for receiving an image captured by the camera and identifying the flame; the device further comprises a pan/tilt used for fixing and adjusting the angle of the camera; the controller comprises a memory and aprocessor as well as a program stored in the memory and executed by the processor; when the processor executes the program, the following steps are performed: aiming the camera at a center point of aflame region after the flame region is identified and obtained; rotating the camera to an another point of the flame region and recording a rotation angle; calculating the number of pixel points between the two points; obtaining the distance between the flame region and the camera based on the number of pixel points, the rotation angle and pre-calibration parameters of the camera. Compared with the prior art, by adopting the device provided by the invention, the dilemma in the prior art that the visual flame recognition can only be simple alarm is overcome by calculating the distance throughtrigonometric functions and the pixel points, and the application range and scene of the flame vision technology can be greatly improved.

The invention discloses a method for quickly decreasing the water-oxygen content of a forming chamber in an additive manufacturing process and a system. As for a sealed atmosphere chamber, a mode of combining vacuumizing, gas washing and circulating purification is adopted, vacuumizing and circulating purification work in a cooperative linkage mode, and thus the water-oxygen content in the sealedchamber can be quickly decreased to the working state; and the preparation time before magnesium alloy/titanium alloy forming is greatly shortened, the working efficiency of equipment is greatly improved, the manufacturing cost of titanium alloy parts is reduced, the production period is shortened, and the socialized income is increased.

The invention relates to a regional exposure algorithm based on weighted gray entropy difference, which is mainly and technically characterized by comprising the following steps of: calculating a weighted gray entropy difference U at each exposure time in a certain fixed scene to obtain a mapping relation between the weighted gray entropy difference and the exposure time; calculating a weighted gray entropy difference of the reference image, and obtaining exposure time by using a dichotomy; dividing the reference image into an over-exposure area Eh, an under-exposure area El and a normal exposure area A; calculating weighted gray entropy differences UA, Uh and U1 of each region; and finally, calculating the optimal exposure time corresponding to each region. The method is reasonable in design, can obtain a good exposure effect in most scenes, has good environment adaptability, can obtain a good imaging effect in complex scenes, and can be widely applied to the field of digital imaging.

The invention discloses a leftover material recycling device for textile production, and relates to the technical field of textiles. The leftover material recycling device for textile production comprises a box body, wherein a driving mechanism, an absorbing mechanism, a processing mechanism, a rotating mechanism and a sweeping and recycling mechanism are arranged on the box body; a connecting plate and two sets of bearing plates are welded and mounted on the outer surface of the bottom of the box body; a set of transmission rods are rotatably mounted on each of the two sets of bearing plates;one end of each transmission rod is rotatably mounted on the outer wall of the connecting plate; and a travelling wheel is welded and mounted at the other end of each transmission rod. The leftover material recycling device for textile production not only can recycle cotton fibers and silk threads floating in air and leftover materials such as rags and broken yarns accumulated on the ground in atextile workshop, but also can clean double surfaces of a textile penetrating through two sets of processing barrels, and recycles thread residues adhered to the textile into a collecting box, so thatthe using range and the using scene are expanded.

The invention discloses a leg of a hydraulic-drive heavy-load six-legged robot, and aims to solve the problems that an existing six-legged robot needs to adapt to terrain and increase the heavy load on the premise of satisfying bionic walking, and that the related research in China is in a blank stage at present. The leg of the hydraulic-drive heavy-load six-legged robot comprises a body connecting plate, a buttock connecting piece, a robot thigh, a robotic shin, a ground supporting piece, a root joint rotating shaft, a horizontal cylinder, a hip joint rotating shaft, a thigh cylinder, a kneejoint rotating shaft and a shin cylinder. The invention provides a completely novel leg structure model. Compared with other existing driving structures, the leg of the hydraulic-drive heavy-load six-legged robot has higher output power and smaller weight in the same space, and has the advantages of higher specific power, lighter weight and higher carrying capacity. Based on the optimized design of the structure of the leg of the hydraulic-drive heavy-load six-legged robot, the terrain adaptability of the six-legged robot is significantly improved. Based on the power matching of a hydraulic system, the leg of the hydraulic-drive heavy-load six-legged robot achieves a breakthrough of 500 kg bearing capacity, and has significant progressive significance.

The invention discloses a method and a device for determining network characteristics of a cell signal path, equipment and a medium. The method comprises the following steps: obtaining a complete network diagram of a specified cell signal path; determining an unmutated gene intersection matched with all the samples according to the corresponding mutated gene set, wherein the mutated gene set is associated with a specified cell signal path; according to the complete network diagram and the mutant gene set of each sample, calculating a non-mutant gene network diagram of each sample, generating anode sequence set of each non-mutant gene network diagram; forming a corpus by using the node sequence set, training to obtain a word vector model for the non-mutant gene intersection, and accordingto a processing result of each word vector model for the non-mutant gene intersection, obtaining cell signal path network characteristics of each sample for a specified cell signal path. The embodiment of the invention provides a brand-new new mode for describing the network characteristics of the cell signal path of the specified cell signal path through the gene mutation data of the sample.

The present invention discloses a vehicle control method and control system based on time-sharing rental. The control method comprises: S1. a mobile terminal sends an unlocking instruction and user account information, a vehicle networking background system pushes the user account information to a vehicle networking terminal system after a vehicle is unlocked, and the vehicle networking terminal system logs into the user account information and updates account data in the user account information; and S2. the mobile terminal sends a locking instruction and user account information, the vehicle networking background system pushes the user account information to the vehicle networking terminal system after a vehicle is locked, and the vehicle networking terminal system logs out of the user account information. According to the method, in the scene of automobile time-sharing rental, the user identity in the vehicle networking terminal system is switched synchronously when a vehicle door is unlocked/locked remotely; and a secure and quick payment operation in the vehicle is realized, in-vehicle user scenarios are extended, and security of user payment is ensured.

The invention discloses a shell deformation resistance detection device for motor production, and relates to the technical field of motor production detection. The shell deformation resistance detection device for motor production comprises a detection table, a pressurizing mechanism, an adjusting detection mechanism, a clamping limiting mechanism and a pressure detection mechanism are arranged onthe detection table, two sets of baffles are welded to the outer surface of the top of the detection table, and a mounting table is welded to the tops of the baffles. According to the shell deformation resistance detection device for motor production, shell deformation is detected in a photoelectric induction mode, the whole operation is simple and convenient, data is accurate and effective, a detection part can be adjusted according to different specifications or forms of motor shells, and the application range and scene of the device are widened. Precise instruments such as professional cameras are not needed in the detection operation of the device, so that the manufacturing cost is greatly reduced, and the popularization and the use are facilitated.

The invention provides a data monitoring method based on a Bluetooth gateway. The method comprises the following steps: setting a Bluetooth gateway and wearable Bluetooth equipment; executing a Bluetooth gateway network access and binding flow; acquiring and parsing broadcast information of the wearable Bluetooth equipment by the Bluetooth gateway; acquiring the broadcast information of the Bluetooth gateway and a video stream image of a camera by a server; intercepting the video frame image according to preset time; acquiring an included angle between a head outline pixel set and the ground,if the angle is less than a preset value, comparing a background pixel set with a ground pixel set, and if the background pixel set matches the ground pixel set, transmitting alarm information to a terminal; and comparing somatosensory information by the server, and if the somatosensory information exceeds a normal value, sounding an alarm and transmitting alarm information to the terminal. Comprehensive analysis is carried out on the broadcast information and video stream information in order to monitor the health status of the elderly in real time.

The invention discloses an antique-finishing treatment device for jean fabric processing of a textile fabric, and relates to the technical field of textile fabric processing. The antique-finishing treatment device for the jean fabric processing for the textile fabric comprises a base; a clamping mechanism, a pushing mechanism, a dust collection mechanism, a treatment mechanism, a driving mechanismand a bearing table are arranged on the base; a mounting frame and a hydraulic rod are welded to the outer surface of the top of the base; a fixing frame is arranged in the mounting frame; and the hydraulic rod is located under the fixing frame, and the free end of the hydraulic rod is welded to the outer surface of the bottom of the fixing frame. According to the antique-finishing treatment device for the jean fabric processing for the textile fabric, the labor burden of an operator is relieved, the rejection rate and the reject ratio are reduced, the product quality is guaranteed, the preparation time required before and after treatment is shortened, more economic benefits are brought, the negative effects on the body health of the operator and the surrounding environment are reduced, and promotion and use are facilitated.