8results about How to "Improve coding accuracy" patented technology

The invention discloses a double-station laser coding device and an operation method thereof. The device comprises a frame, a first coding platform mechanism, a second coding platform mechanism, a laser coding mechanism, a first code scanning platform mechanism, a second code scanning platform mechanism, a code scanning mechanism, a material moving mechanism, a product feeding mechanism and a product receiving frame. The invention adopts two groups of coding platform mechanisms, and is equipped with two groups of lasers and CCD galvanometer coaxial components to form double-station laser coding, which are independently controlled and do not affect each other, and are equivalent to integrating two single-station laser coding devices, thereby greatly improving the coding efficiency. Two groups of code scanning platform mechanisms are adopted to form double-station code scanning, and a common set of scanning mechanisms are used. The double stations alternately operate to scan codes, and the working time of the scanning mechanisms is fully utilized, thereby greatly improving the scanning efficiency. The invention has both coding and scanning functions, and integrates coding and scanning on one device, thereby reducing a scanning process, a set of scanning equipment and labor, thereby improving production efficiency and reducing costs.

The utility model discloses a kind of hollow punch coders of adjustable depth of steel plate surface, including coding base and clamping mechanism, coding base: its upper end front side is equipped with bottom plate, the upper end of bottom plate is equipped with liftable clamping plate, clamping mechanism: it includes moving block, limit slot, connecting rod, clamping assembly and lifting assembly, the moving block is slidably connected in the left and right sides inside clamping plate respectively, limit slot is all set in the inside of moving block close to the side in the middle of clamping plate, connecting rod is all rotationally connected in the inside middle of clamping plate, this adjustable depth hollow punch coder of steel plate surface, while fixed steel plate can also form limiting effect to steel plate surface, and make steel plate below and clamp bottom wall separate, reduce the influence of laser on clamp, steel plate can also be moved slowly to ensure that the interval between the surface of the coding place of steel plate and laser between in coding process is always equal, reduce the vibration of laser, improve the coding accuracy of laser.

The present application belongs to the field of computer technology, and specifically relates to a training method for a summary extraction model, a text summary generation method and related equipment. Embodiments of the present invention can be applied to various scenarios such as cloud technology, artificial intelligence, smart transportation, and assisted driving. The method includes: obtaining a first loss function for training a summary extraction model; periodically updating the training parameters of an auxiliary encoder according to the training parameters of the encoder; calculating a first similarity between a text encoding of a text sample and a summary encoding of a corresponding summary sample, calculating a second similarity between the text encoding of a text sample and summary encodings of other summary samples, and determining a second loss function according to the first similarity and the second similarity; and training the summary extraction model together through the first loss function and the second loss function. Based on the present application, the accuracy of text summary generation of the summary extraction model can be improved, so that the generated text summary is more consistent with the semantics of the original text to be processed.

The embodiment of the present invention discloses an image encoding method, device, computer equipment and storage medium. The method comprises: in at least one standard encoder type, detecting the supported encoder type of the current device, wherein the supported encoder type at least comprises a standard software encoder type; selecting an encoder according to each of the supported encoder types, and encoding the image data to be encoded. The embodiment of the present invention can be compatible with a hardware-free encoding environment and improve the availability of the encoding system.

The utility model discloses a coding machine for packaging food auxiliary materials, which comprises a mounting frame arranged above a machine base, material conveying rollers are arranged on the inner side of the mounting frame, a conveying belt is arranged between the two material conveying rollers, and a stand column is arranged on one side of the conveying belt; the transverse plate is arranged on one side of the stand column, a first sliding groove is formed in the side wall of the stand column, a first sliding block is arranged in the first sliding groove, a movable plate is arranged below the transverse plate, a second sliding groove is formed in the lower end of the transverse plate, and a second sliding block is arranged in the second sliding groove; a laser code printing mechanism is arranged below the moving plate, a third sliding groove is formed in the lower end of the moving plate, a third sliding block is arranged in the third sliding groove, the laser code printing mechanism can move transversely or longitudinally, different patterns are printed according to code printing requirements, and food packaging and conveying are seen through the moving conveying belt; and the food packaging coding efficiency is improved on the premise of ensuring the coding precision.

This invention relates to the field of QR code technology, specifically to a character code generation method, decoding method, and apparatus. The method includes: acquiring text to be encoded and converting it into information to be encoded through an encoding mode; creating a scanning area and filling the scanning area with target text, wherein the version of the scanning area is determined based on the size of the target text and the size of the information to be encoded; splitting the information to be encoded into one or more data modules; matching candidate filling positions for each data module by traversing the outline formed by the strokes of the target text according to the size of the split data modules; and filling each data module into its corresponding filling position after a successful match. This invention provides a QR code encoding and decoding method that can be recognized by both machines and the human eye. Each character has a fixed outline path formed by strokes, and by changing the control points on the outline to embed encoding information, machine recognition and human eye recognition are combined.

A bitstream is received, comprising attribute information of a mesh having multiple vertices. It is determined whether a parallelogram-in-place prediction pattern is applicable to the current vertex among the multiple vertices of the mesh. When it is determined that the parallelogram-in-place prediction pattern is applicable to the current vertex of the mesh, one or more parallelogram-in-place predictions for the current vertex of the mesh are generated. Based on (i) the one or more parallelogram-in-place predictions, and (ii) one of the prediction residuals and index syntax included in the bitstream, the current vertex of the mesh is reconstructed.

This invention discloses a method and system for intelligent coding review and automated data entry of medical records, relating to the field of data processing technology. The method acquires structured data from the electronic medical record system, including the patient's discharge diagnosis, treatment process, pathology report, medication records, and cost information, through multi-source acquisition. The data is then classified and preprocessed according to medical record type. A pre-trained natural language processing model analyzes the text data in the medical record homepage to understand the doctor's diagnostic descriptions and treatment process. Combined with a constructed coding rule base and knowledge graph, a recommended coding list and anomaly alerts are output through rule mapping and reasoning. A human-machine collaboration mechanism is established to automatically enter the reviewed medical record coding data into the medical record management system. This solves the problems of excessive reliance on manual operation, insufficient coding accuracy, and low efficiency in existing medical record coding review methods, forming a precise medical record coding working mode based on human-machine collaboration.