12results about How to "Efficient and accurate classification" patented technology

The invention relates to the technical field of image recognition, in particular to a self-adaptive image classification method and system for a medical image data set. The method comprises the following steps that a to-be-classified medical image data set is obtained, a key interest area of each image is extracted, and the to-be-classified medical image data set comprises a plurality of X-ray medical images and CT medical images; performing frequency domain conversion on the to-be-classified medical image data set according to the key region of interest, and respectively analyzing the texture fine granularity of each image in the to-be-classified medical image data set to obtain high-frequency texture fine granularity and low-frequency texture fine granularity; and integrating the high-frequency texture fine granularity and the low-frequency texture fine granularity as first classification feature points. According to the invention, through adaptive image enhancement, texture fine-grained analysis and contrast difference calculation, the accuracy and adaptability of classification of different types of medical images are improved.

The application provides a ConvNeXt model SAR ship classification method based on multi-level feature cooperative interaction. In order to effectively utilize traditional manual features, the method uses a Canny-based edge detection method, captures edge information of objects in an image through multi-level filtering and edge gradient analysis, and inputs the result into a ConvNeXt model for feature extraction. In addition, a multi-level branch structure is expanded in the ConvNeXt model, which is used for deep extraction of feature information at different scales, so as to capture various information of the image at low-level features and high-level features. Finally, cross fusion with traditional manual features is carried out in the multi-scale feature extraction process, so as to improve the classification accuracy of the model.

The invention relates to the technical field of image data processing, in particular to a milk powder category detection method and system based on image processing. Calculating the character calibration importance degree of each feature point to represent the deviation degree of the feature points relative to the character region; calculating the character possibility degree of the feature points, and synthesizing the probability that the feature points belong to the character region in the image; defining the product of the character calibration importance degree and the character possible degree as the calibration necessity degree of the feature points; according to the method, feature points with calibration necessity higher than a set threshold value are screened out, a homography matrix is calculated through the feature points, character areas are extracted to recognize character information after perspective transformation, and the recognized character information is input into a classification model to complete automatic classification of milk powder categories. The problem that the accuracy is insufficient in the milk powder classification process is solved.

The invention discloses a self-adaptive data lossless acquisition and storage method, which comprises the following steps of performing data feature extraction on existing data in a database, and associating data features with a storage medium; data feature capture is performed on data needing to be collected, and a medium suitable for storage is analyzed according to data features; in the process of data acquisition, transmission and storage, the data are verified; according to the self-adaptive data lossless collection and storage method, the storage medium is automatically analyzed through data feature analysis, a data source does not need to be manually configured, the data is verified under various operations, the integrity and consistency of the data are ensured, the data are detected, and the data recovery efficiency is improved. The efficiency of timely repairing the data is ensured, and the accuracy and value of the data are ensured.

The invention belongs to the field of information processing, and provides an information processing method and system for voice conversion, and the method comprises the steps: obtaining all voice signal data needing voice conversion based on all interfaces for collecting voice data, carrying out the noise reduction processing, and obtaining an initial noise reduction voice signal; acquiring fundamental frequency and formant positions of the voice signals, analyzing and processing the fundamental frequency and formant positions to obtain gender characteristic values, and dividing the gender characteristic values; based on the voice information subjected to gender classification, waveform data of the voice information is acquired and subjected to comprehensive analysis processing to obtain emotion feature values, and classification processing is carried out; on the basis of the voice information after noise reduction processing, official text information corresponding to the voice information is obtained through conversion, and gender and emotion information carried by the voice information is added to the text information; according to the technical scheme, official language characters obtained after voice conversion are converted into official voice information, gender and emotion information carried by the voice information is attached to the converted voice information, the man-machine interaction quality is improved, and the human daily life and work experience is improved.

The invention relates to the technical field of enterprise digital management, in particular to an enterprise management system and method based on data intelligent analysis processing, and the system comprises a data collection module, a data processing module, a data analysis module, a data storage module, a data visualization module and a decision support module. According to the system, the data analysis module is arranged, the data analysis module carries out deep mining on various types of data such as documents, numerical values, image videos and audios through refined classification and a targeted analysis strategy, and advanced technologies such as text processing, feature extraction, video framing and a 3D convolutional neural network are utilized to carry out deep mining on various types of data such as the documents, the numerical values, the image videos and the audios; according to the method, efficient and accurate data processing and classification are realized, and finally, various analysis results are integrated, so that comprehensive and deep data support is provided for enterprises, the decision-making process is assisted to be optimized, and the business insight and the operation efficiency are remarkably improved.

The invention relates to the field of data processing, in particular to a deep learning litigation material preprocessing and classifying method. The method comprises the following steps: acquiring initial litigation material data in a database; converting the initial litigation material data into a high-dimensional vector by using a pre-trained Word2Vec word embedding model to obtain coded litigation material data; performing feature extraction on the coded litigation material data through a CNN convolutional neural network to obtain feature litigation material data; establishing a B i-LSTM litigation material classification and prediction model based on a bidirectional B i-LSTM long-short term memory network, and performing combinatorial optimization on the B i-LSTM litigation material classification and prediction model by using a whale optimization algorithm; and inputting the feature litigation material data into the target B i-LSTM litigation material classification prediction model for classification, obtaining a user query record of encrypted and classified litigation material data in real time, and performing data query recommendation classification according to the user query record. The category to which the litigation material belongs can be automatically judged according to the input litigation material, and efficient and accurate classification is realized.

This invention relates to the field of unmanned aerial vehicle (UAV) inspection technology, specifically an intelligent analysis system for UAV river and lake inspection in smart water conservancy. The system includes a UAV equipped with visible light and thermal infrared sensors performing a dual-temporal detection mission to capture images and data under different lighting and heat conditions. In the anomaly identification stage, a generative neural network trained on normal river and lake images is used to initially locate anomalies by calculating the error between real-time images and model-reconstructed images. A water surface reflection suppression network is introduced for secondary processing to filter out specular reflection artifacts and generate candidate anomaly areas. Anomaly diagnosis employs a two-level strategy: first, a rapid coarse classification is performed by calculating apparent thermal inertia values ​​based on dual-temporal thermal infrared data; then, a secondary fine-tuning diagnosis is performed by combining depth texture and higher-order spectral moment analysis to achieve accurate classification of anomaly types. This invention enables precise and automated identification and early warning of river and lake anomalies through UAV inspection.

The application relates to an automatic classified biomass garbage recycling treatment system for a zero-carbon park, which comprises a garbage classification system, a garbage treatment system, a storage system and a cleaning system. The garbage classification system classifies and processes garbage, the garbage treatment system recycles and processes garbage, the storage system stores the recycled and processed garbage, and the cleaning system cleans the garbage classification system. The automatic classified biomass garbage recycling treatment system for the zero-carbon park can solve the problems that garbage classification often depends on manual work or single technology, the classification accuracy is low, a large amount of garbage is still landfilled or incinerated, energy is wasted, carbon emission is caused, especially the biomass waste such as peels and fallen leaves generated in the park, which needs to be collected and transported to the outside for traditional treatment, thereby increasing traffic energy consumption and carbon footprint.

The present invention provides a method, system, computer device and storage medium for identifying pig attack behaviors. The method comprises the following steps: extracting a sequence of key frames to be identified from a video of a group of pigs to be identified, inputting the video of the group of pigs to be identified into a lightweight 3D convolutional neural network to extract corresponding spatiotemporal features to be fused, and inputting the sequence of key frames to be identified into a 2D convolutional neural network to extract corresponding spatial features to be fused, then fusing the spatiotemporal features to be fused with the spatial features to be fused to obtain spatiotemporal fusion features, and using a hole convolution layer to perform bounding box prediction and action classification on the spatiotemporal fusion features to obtain an identification result. The method solves the application limitations of the prior art, such as excessive reliance on human experience, time-consuming and labor-intensive, and insufficient intelligence, while achieving accurate and efficient identification and classification of pig actions, providing real-time and effective monitoring of pig attack behaviors, and providing reliable guarantees for the healthy growth of pigs and the economic income of the pig farming industry.

The application provides a fabric pilling automatic grading method and device, and relates to the technical field of textile detection. The method comprises the following steps: collecting a thermal infrared polarization image sequence of a fabric sample by using a thermal infrared polarization camera; calculating a polarization degree and a thermal decay time constant based on the sequence data, and extracting a micro-thermal radiation feature; simultaneously collecting a visible light image by using a visible light camera, and extracting an image feature; fusing the micro-thermal radiation feature and the image feature to generate multi-modal feature data; and finally identifying and determining a pilling grade by using a neural network model. By combining thermal physical characteristics and visual features, the problems of strong subjectivity of artificial grading and large limitation of single optical detection are overcome, and objective, accurate and efficient automatic grading is realized.

The application discloses a virus-host RNA sequence classification method and device based on AI, and relates to the field of biological detection.The method comprises the following steps: mapping preprocessed short read RNA sequences to a host genome twice, assembling short read RNA sequences which are not mapped to the host genome into continuous RNA sequences, and screening RNA sequences with a length greater than 1000bp from the continuous RNA sequences; and performing AI classification on the RNA sequences with a length greater than 1000bp to obtain virus RNA sequences.The application combines host filtering, rapid assembly and AI classification, can significantly reduce the calculation amount and hardware pressure, realizes efficient and accurate virus sequence classification, and can quickly distinguish unknown viruses.