14results about How to "Improve automated processing efficiency" patented technology

The invention discloses a multi-modal document structured processing and knowledge extraction method based on a large language model. The method comprises the following steps: S1, receiving an input multi-modal document; s2, performing preprocessing on the text data in the multi-mode document; s3, performing feature extraction on non-text data in the multi-mode document; s4, performing multi-modal data fusion on the preprocessed text data and the non-text data after feature extraction; s5, performing deep semantic analysis on the fused multi-modal data through a pre-trained improved BERT model, wherein the deep semantic analysis comprises key entity recognition, relation extraction and event detection; s6, based on a deep semantic analysis result, automatically constructing a knowledge graph by using the extracted information; and S7, outputting the data of the knowledge graph in a format for analysis or application. According to the method and the system, structured processing and knowledge extraction of unstructured and multi-modal documents are realized.

The invention discloses an unsupervised multi-stage large geometric deformation remote sensing image initial registration method and model, and the method comprises the steps: 1), building a pre-training data set, which comprises the construction of the pre-training data set and the construction of an aerospace image fine-tuning data set; 2) constructing a large geometric deformation remote sensing image registration model, designing the model into four independent parameter estimation modules, and gradually recovering an approximate geometric transformation relationship among large geometric deformation remote sensing images through conversion parameter calculation design in multiple stages; and 3) training a large geometric deformation remote sensing image registration model, independently pre-training the four parameter estimation modules by using the constructed pre-training data set, and uniformly training the whole model by using the aerospace image fine tuning data set. According to the method, the division and treatment strategy is adopted, the mode of module independent training and then model unified training is adopted, the image registration success rate is greatly increased, and the method has remarkable engineering application value.

The utility model belongs to the technical field of bending machines, and particularly relates to an actuating mechanism of a bending machine, which comprises an operation table and a negative pressure pipe, a bending table is arranged at the top of the operation table, a bending machine body is arranged above the bending table, an actuating piece is arranged at the top of the bending machine body, and a bending mechanism is arranged on the negative pressure pipe. A bending pressing block is arranged on the bending machine body, the negative pressure pipe is arranged on the front side wall of the bending pressing block, and an adsorption cavity is formed in the bending pressing block. In the bending machine, an automatic execution material taking mechanism is additionally arranged on the bending machine, so that the automation degree of equipment is greatly improved, automatic material taking after antenna auxiliary materials are bent is achieved, and the working efficiency is improved. Therefore, the automatic processing efficiency of the equipment is improved.

The invention relates to a board placing device and a control method thereof, and relates to the technical field of PCB automatic machining, the board placing device comprises a protection cover, a controller installed outside the protection cover and a rack installed in the protection cover, the rack is provided with a feeding table close to the feeding end, and the rack is provided with a discharging table close to the discharging end; an anti-sticking mechanism is arranged on the feeding table, the anti-sticking mechanism comprises first moving equipment, a plurality of fixed anti-sticking assemblies and a plurality of movable anti-sticking assemblies, and the first moving equipment is used for driving the movable anti-sticking assemblies to move in the direction close to or away from the fixed anti-sticking assemblies; a thickness detection mechanism is mounted above the discharging end; a moving mechanism is arranged above the feeding table and the discharging table and comprises second moving equipment and adsorption equipment, and the second moving equipment is used for driving the adsorption equipment to move in the direction close to or away from the feeding table or the discharging table. The method has the effect of improving the automatic processing efficiency of the flexible PCB.

The invention relates to the technical field of image analysis, in particular to a processing method for a natural gas pipeline threaded interface, which comprises the following steps: acquiring a plurality of threaded edge lines, non-threaded edge lines and threading die edge contours in a processing image of the natural gas pipeline threaded interface at the current moment, further determining the thread edge ambiguity in the processing image and the integrating degree of the threading die and the thread in the processing image; according to the integrating degree of each moment in the preset monitoring time period and the vibration monitoring data of the processing equipment, determining the fuzzy confidence coefficient of the thread edge; and determining a scale factor of Wiener filtering by combining the thread edge ambiguity, the integrating degree and the thread edge ambiguity confidence coefficient at the current moment, performing filtering processing to obtain a processed processing image, and extracting image features to dynamically adjust processing parameters at the current moment. According to the method, the filtering parameters during filtering processing of the image in the threaded connector machining process are determined in a self-adaptive mode, and the machining quality of the threaded connector is effectively improved.

The present invention relates to the technical field of tube laser cutting, and in particular to a dual-axis synchronous laser cutting machine suitable for plastic tube processing and a control system thereof. The present invention comprises a supporting base, the upper surface of the supporting base is fixedly connected to a processing table, the front surface of the processing table is fixedly connected to a control panel, and one side of the processing table is symmetrically fixedly connected to a servo motor; the present invention completes the precise cutting of plastic tubes by mechanical and system linkage, which helps to improve the cutting efficiency of plastic tubes, that is, the flip wheel is mechanically driven to drive the plastic tube to rotate so as to achieve overall cutting, and through image data analysis at both ends of the plastic tube, it helps to achieve automatic identification, positioning and cutting of both ends of the plastic tube, and by calibrating, tracking and supervising the cutting data, the cutting accuracy of the equipment and the cutting quality of the plastic tube are improved.

This application relates to a fully automatic keyboard assembly and key replacement machine, comprising two replenishment machine bodies and a movable machine body. The movable machine body is equipped with a moving table and a riveting assembly for riveting keyplates onto a base plate. A temporary storage platform is provided between the replenishment machine body and the movable machine body. A suction head is provided within the replenishment machine body, along with several vibrating and flat vibrating hoppers. A correction assembly is located between the vibrating and flat vibrating hoppers, including a feeding channel installed between them. The feeding channel has suction and leakage notches. A rotating cylinder is rotatably connected to the feeding channel, and several suction nozzles are provided on the rotating cylinder. A rotating assembly is provided on the feeding channel, and a blocking plate is provided on the suction notch. A receiving box is detachably connected to the replenishment machine body. A vibrating component assisting in keyplate movement is provided on the feeding channel, and a vision camera is located at the top of the replenishment machine body. This application improves the automation efficiency of the key replacement machine.

The invention discloses an animation video frame intelligent selection dialogue generation system, which specifically relates to the field of computer vision technology. The animation video in a character library is converted into a video frame sequence, and the character appearing in each frame is detected and tracked using computer vision technology. For the detected character, its appearance features are extracted, and the extracted character appearance features are matched with the character appearance features obtained in the character library to determine the identity of the detected character. A character timbre library is established, including audio data and annotated timbre. Audio features are extracted from the audio data and timbre recognition is performed. The recognized timbre is matched with the pre-annotated character timbre. According to the timbre recognition result, the dialogue content of each character is extracted, and the dialogue content is converted into text form. The extracted dialogue content is judged for topic similarity. According to the judgment result, the dialogue content with similar topics is modified to generate new dialogue content.

The application relates to the field of metal flaw detection, in particular to a conveying type pipe flaw detection equipment, which comprises a rack, a conveying belt installed on the rack, the rack is provided with a magnetization flaw detection station and an observation station, the conveying belt can drive the pipes to pass through the magnetization flaw detection station and the observation station in sequence; a magnetization flaw detection mechanism, which comprises a magnetization coil connected to the rack, the magnetization coil is used for magnetizing the pipes moving to the magnetization flaw detection station along the conveying belt; the rack is further connected with a magnetic powder pipe, the magnetic powder pipe is connected with a magnetic powder spray head used for spraying the magnetic powder to the surface of the steel pipe; an observation mechanism, which comprises a plurality of industrial cameras, the industrial cameras are used for shooting and recording the surface defect magnetic traces of the pipes, the observation rack is further connected with a driving element and a marking rack, the marking rack is connected with a plurality of spray pens, the driving element can drive the marking rack to move to the direction of the pipes in the observation station, and the spray pens can spray and draw marks on the surface of the pipes. The application has the effect of detecting the surface defects of the metal pipes by automatically spraying the magnetic powder to the surface of the pipes and accelerating the inspection speed.

The invention relates to a stacking device and a control method thereof, and relates to the technical field of automatic machining, the stacking device comprises a stacking platform, a mounting box, a grabbing mechanism, a lifting mechanism and a controller, and the lifting mechanism is used for driving the grabbing mechanism to ascend and descend; the grabbing mechanism comprises a robot arm device, a plurality of position feedback devices and a plurality of state detection devices, the position feedback devices and the state detection devices are installed on all parts of the robot arm device, and the position feedback devices are used for feeding back real-time position data of all the parts of the robot arm device. The state detection equipment is used for detecting real-time state data of each part of the robot arm equipment; and the controller is used for updating a planned path according to the real-time feedback data and the real-time state data, and controlling the grabbing mechanism and the lifting mechanism to work according to the updated planned path. The automatic machining device has the effect of effectively improving the automatic machining efficiency.

The invention discloses an automatic task processing method and device, and the method comprises the steps: receiving a to-be-processed task uploaded by each user, and generating a plurality of task queues based on the to-be-processed task; determining estimated resource occupation information of the to-be-processed task, and adjusting a current node cluster based on the estimated resource occupation information to obtain a target node cluster; distributing the to-be-processed tasks in the task queue to each node in the target node cluster according to a predetermined sequence; and after each node finishes processing the to-be-processed task, returning a task processing result to a user, and dynamically optimizing the distribution sequence of the to-be-processed task and the node scale of the current node cluster by comparing the processing capabilities of the to-be-processed task and the current node cluster so as to improve the efficiency of automatic task processing.

The utility model discloses inner cavity polishing equipment for processing an American bucket with an engine oil nozzle. The inner cavity polishing equipment comprises a bottom plate, a material conveying mechanism, a fixing mechanism, a polishing mechanism and a cleaning mechanism, wherein the material conveying mechanism, the fixing mechanism, the polishing mechanism and the cleaning mechanism are arranged on the bottom plate. Through cooperation of the material conveying mechanism, the fixing mechanism and the polishing mechanism, the American belt machine nozzle barrel is automatically conveyed to a designated position, then the American belt machine nozzle barrel is fixed through the fixing mechanism, a polisher can accurately enter an inner cavity of the American belt machine nozzle barrel, and the polishing efficiency is improved. And finally, the polishing work of the American belt machine nozzle barrel is completed through the polishing mechanism, the step of correcting the position of the American belt machine nozzle barrel midway is omitted, and the automatic machining efficiency is improved.

The utility model provides a kind of printing machinery rapid cleaning device, it is related to printing machinery technical field, including in and out material output assembly and clamping turnover parts, the middle part inboard of in and out material output assembly is provided with spray washing mechanism, one end of in and out material output assembly is provided with clamping turnover parts above, the clamping turnover parts include lower open box, transverse screw rod group, sliding base, grooved frame, lifting screw rod group, lifting block, electric damping shaft, turnover cross seat and electric clamp, the lower open box is set in one end of in and out material output assembly above;The utility model mainly is output operation using the multiple groups of same direction guide pulley of parallel distribution, so that the lower open box, transverse screw rod group, sliding base, grooved frame, lifting screw rod group, lifting block, electric damping shaft, turnover cross seat and electric clamp are clamped turnover and send out next layer for the material of upper layer output under the cooperation, to improve the automated processing efficiency of equipment.