8results about How to "Solve classification problems" patented technology

The invention discloses an anti-shake method for an image transmission video of an unmanned aerial vehicle. The anti-shake method comprises the following steps: step 1, data acquisition; step 2, multi-frame optical flow analysis; step 3, feature extraction and jitter judgment; step 4, generating a jitter-free frame by the generative adversarial network; and step 5, post-treatment and synthesis. And step 6, performing real-time feedback and optimization. Through combination of multi-frame optical flow analysis and deep learning feature extraction, different types of jitters can be accurately identified. The accurate jitter judgment not only improves the reliability of detection, but also provides an important basis for subsequent processing, so that the video quality can be effectively guaranteed in practical application no matter in a complex environment or in a rapid movement condition, and the scheme can remarkably improve the video quality on the premise of keeping details and definition. The visual effect of the image is further enhanced through subsequent color correction and sharpening processing, so that the finally output video is smooth and has higher watching value.

The utility model discloses a desktop formula test tube labeling machine, including buffer memory device, handling device, labeling device, tray storage device and tray transmission device, buffer memory device is used for buffering test tube, and buffer memory device is equipped with hopper, and the aperture of hopper's mouth is greater than the outer diameter of test tube's pipe cap, and the minimum aperture of hopper is slightly greater than the outer diameter of test tube's pipe cap, handling device is used for handling test tube located in buffer memory device to hopper, the entrance of labeling device is located just below hopper, and labeling device is used for test tube to label, tray storage device can supply tray and storage tray, tray transmission device, tray transmission device can transmit the tray that tray storage device supplies to just below the export of labeling device, and can transmit the tray located just below the export of labeling device to tray storage device and store. It can solve the problem that the existing test tube labeling machine cannot arrange the pipe by oneself, and also cannot classify and store the test tube that labels are completed.

The embodiment of the present application provides an image classification method and device, the method comprising: obtaining an object image to be classified, a foreground target extractor and a final classifier; inputting the object image into the foreground target extractor to obtain a foreground target image; inputting the foreground target image into the final classifier to obtain a classification result; the final classifier is composed of an image feature extractor and a small sample object classifier; the image feature extractor is obtained by training with a large sample of foreground target images of known types of objects as training data; the small sample object classifier is obtained by training with an expanded foreground target image of a new type of object as training data; the expansion method comprises: using a posture conversion generator to convert the foreground target image of the new type of object into an expanded foreground target image. The embodiment scheme effectively solves the classification problem of fine-grained images under small sample conditions and effectively improves the classification accuracy.

This invention discloses a novel detector method based on a two-stage detection model. This invention has certain versatility within two-stage detection models, and the improved detection model can also be used as a standalone detection framework. To address the problem that traditional two-stage detection models often misclassify new targets as easily confused base class targets in classification tasks, this invention adds a contrastive network to the detector to understand intra-class similarities and inter-class differences. By contrasting object encoding, inter-class errors can be reduced. The patent verifies the feasibility and usability of the algorithm through ablation and contrast experiments, thereby demonstrating the feasibility and rationality of the novel detector proposed in this invention.

The invention relates to the technical field of book management, in particular to a library book sorting system based on image recognition, which comprises an image acquisition module; an image processing module; the interdisciplinary association analysis module is used for analyzing the content association between the interdisciplinary books after the preliminary book sorting, and the interdisciplinary association analysis module is used for analyzing the content association between the interdisciplinary books; the knowledge path recommendation module is used for automatically identifying the same knowledge path book based on an output result of the interdisciplinary association analysis module and generating an optimized virtual knowledge path sorting scheme; providing an interactive interface for a borrower to check and select a highlighted knowledge path recommendation book; and the data storage module is used for storing the images, the classification labels, the associated information and the borrowing historical records of the books. According to the method and the device, the books can be optimized and sorted according to the content relevance and the requirements of the readers, so that the efficiency of searching the relevant books by the readers is remarkably improved, and the problem that the book classification and recommendation are not accurate enough in the prior art is solved.

The present invention discloses a small sample scene image classification method based on a multi-graph guided neural network model. Based on the features of the multi-graph guided neural network model at different levels (view), image feature information at different levels is modeled, and a graph convolutional neural network for relationship modeling of different samples is introduced at the same time. The sample features are pulled closer to the same type of samples and away from different types of samples through feature propagation between samples, so that the sample features are more discriminative. The present invention can effectively solve the problem of scene image classification with small samples and complexity. Its model architecture mainly includes four aspects: using a deep network to extract features of different levels of views of the scene image, constructing a graph based on the global feature representation of views at different levels, performing graph convolution representation of features of views at different levels respectively, and classifying query samples based on the features guided by multiple graphs.

The invention discloses a factoring business management method based on a block chain. The method comprises the following steps: S1, a management system on a factoring business chain obtains related data of a factoring business participant; s2, the factoring service chain management system classifies and records the data; s3, the factoring service on-chain management system calls a regional chain network to link the data and obtain a link-up result, and if link-up succeeds, the data is cached; s4, performing factoring service application in the factoring service system, calling a factoring service on-chain management system, performing on-chain query and calling of data required by the application according to query conditions, returning the data to the factoring service on-chain management system, and returning the data to the factoring service system by the factoring service on-chain management system; and S5, the factoring service system audits the application, and performs data right confirmation or performs the next service on the basis of the on-chain data. The method has the advantages of factoring service data classification, data right confirmation, data use record traceability and the like.

This utility model relates to the technical field of deoiling systems and provides a waste catalyst deoiling system. It includes a deoiling furnace, a slag remover, a bucket elevator, a screen, a hopper, and an economizer. The deoiling furnace includes a moving block and a grate. The grate is composed of a crossbeam and grate plates, with the grate plates evenly assembled on the crossbeam. The crossbeam is fixedly connected to the moving block, and pulleys are mounted on the moving block. The rotation of the pulleys causes displacement of the moving block and the grate. The slag remover and the bucket elevator cooperate to feed the material discharged from the slag remover into the screen. The discharge port of the screen connects to the inlet of the hopper. The flue gas outlet of the deoiling furnace is provided with a flue that connects to the economizer. This waste catalyst deoiling system can distribute the force on the grate through the moving block, reducing the grate deformation and lowering the possibility of problems such as material leakage and chain slippage.