32results about How to "Easy to classify and identify" patented technology

The invention discloses a low-rank partitioning sparse representation human face identifying method, which adopts low-rank matrix decomposition, introduces a reference item, adopts a DCT (discrete cosine transform) algorithm to realize the normalization of images and effectively solves the problem of uneven lighting in a human face image. At a classifying stage, a clustering thought is used, and the identifying speed is effectively improved. The algorithm is used on a standard human face database to perform multiple times of tests, and the test result shows that compared with the existing human face identifying method, the low-rank partitioning sparse representation human dace identifying method has the advantage that the identifying accuracy and computing efficiency of the algorithm are all consistently improved. The precision and stability on human face identifying are improved under the complex conditions, such as shielding, disguising and illumination varying, of the human face image.

The invention discloses a train rolling bearing fault and impurity vibration distinguishing method. The method includes: utilizing spectral kurtosis to perform filtering processing on an original bearing vibration signal, and adopting an empirical mode decomposition technique to perform step-by-step decomposition on the preprocessed vibration signal to obtain intrinsic mode functions; subjecting the intrinsic mode functions to frequency spectrum and statistic characteristic analysis, extracting a feature vector composed of frequency spectrum center-of-gravity and characteristic frequency energy ratio as input, establishing a support vector machine based classifier model by training, and distinguishing whether a rolling bearing with abnormal noise is faulted or contains impurities in real time. The method is used for distinguishing the fault or impurity-containing state of the train rolling bearing based on vibration signal analysis, the method effectively improves accuracy in bearing fault identification, the problem of overhigh false alarm rate in the bearing vibration diagnosis method caused by impurities is solved, and good technical support is provided for engineering application of the vibration diagnosis method.

The invention relates to a magnetic shoe surface defect detection system and method, and relates to the field of magnetic shoe surface defect detection. The system comprises an image acquisition device and a computer system. The image acquisition device is used for acquiring magnetic shoe surface images. The computer system performs suspicious area judgment on the magnetic shoe surface image and divides the suspicious area to acquire a suspicious area image. The computer system extracts geometric features, gray features and texture features of the suspicious area image. The computer system inputs the geometric features, the gray features and the texture features into a classifier which is trained in advance. The detection result of the magnetic shoe surface defects is obtained, high-precision classification and identification of the magnetic shoe defects are realized through a classification and identification algorithm, automatic detection of the magnetic shoe defects is realized, thelabor cost is reduced, and the efficiency and the identification accuracy are improved.

The invention discloses a tobacco processing logistic turnover box, which comprises a base (4), a box body (1) and a box cover (6). The invention is characterized in that the box body (1) and the box cover (6) are matched in a buckle connection type, the bottom of the box body (1) is fixedly connected on a base (4) provided with a forking groove, and an identifiable card (5) is arranged at the side surface of the base (4) or the box body (1). The invention adopts a standardized logistic turnover box, is convenient for the classified identification of materials in each process flow of the tobacco processing, is also convenient for the turnover of a fork truck or a piling machine, greatly improves the work efficiency, and reduces the production cost. Reinforcing ribs are arranged on the box body, the integral structural intensity of the box body is improved, and a force exertion point is also provided for automatic box turning equipment, so the adaptability of the turnover box to the process flow is enhanced. The tobacco processing logistic turnover box can be made of wood materials, and the occurrence of the phenomenon that the stainless steel turnover box cover in the prior art easily generates condensed water beads or generates static electricity is avoided. The invention has the advantages of simple and convenient manufacture and flexible dispatching, the base with easy consumption in the use process can be conveniently replaced, and the invention is suitable for the large-scale tobacco processing process.

The invention relates to a filtering and intercepting method for junk mail. The method filters and recognizes the mail for multiple layers: obtaining the classification attribute of mail for the user according to the expression and/or phonetic feature data of the user for receiving and sending the mail; for the mail, which cannot be confirmed on category, orderly inquiring the black/white list of the user node local and friend node, the friend node is the network node having higher mail interaction frequency with the current user node; reminding the user for judging the category of the mail. By filtering and recognizing the mail for multiple layers, the problem of the prior art for recognizing the junk mail with larger cost and low efficiency can be solved.

The invention provides a multi-label identification algorithm based on DCGAN and GCN, and the algorithm comprises the steps: building a DCGAN model based on a GAN model, and generating a similar image based on the DCGAN model; extracting features based on a migrated CNN algorithm, migrating parameters of a neural network of the DCGAN model to the CNN algorithm to perform feature extraction on the multi-label image, and generating a category label classifier through a relation graph between training labels by using a GCN algorithm; generating a data pre-training model through a deep convolutional generative adversarial network, and migrating parameters of a convolutional neural network of the pre-training model to a target task to finely adjust the network so as to obtain a more accurate image identification effect. Meanwhile, when the image is generated, random noise is increased, so that the robustness of the pre-training model can be improved.

The invention discloses a radar radiation source identification system based on signal scale decomposition. The system is composed of a radar signal database, a radar receiver, a signal decompositionmodule, a classifier and an identification result display instrument. The radar signal database stores historical collected radar signals of known types, and the radar receiver receives latest radar signals of unknown types. All radar signals are subjected to data multi-scale decomposition processing through a signal decomposition module, then the type of a radar radiation source is identified bya classifier according to the processed data, and finally an identification result display instrument displays a result. According to the method, the intra-pulse features of the radar radiation sourceare extracted through multi-scale decomposition and the lightweight gradient boosting tree, and classification and recognition can be better carried out compared with conventional features. When theradiation source signal is subjected to deep multi-scale analysis, the signal is subjected to multi-frequency decomposition, and in order to prevent frequency confusion, the original signal is subjected to processing of adding white noise and averaging, so that the identification accuracy is further improved.

The invention provides a fly face recognition method based on a deep convolutional neural network. The method comprises the steps of optimizing a multi-task convolutional neural network MTCNN to perform face positioning and feature point detection on a fly image; obtaininig a fly face feature point image through positioning and detection, and carrying out fly face alignment; making the aligned flyface images into a data set; building a fly face deep convolutional neural network model; connecting two groups of feature vectors extracted from the thickness in the fly face deep convolutional neural network in series to form a group of feature vectors; and testing the fly face deep convolutional neural network model through the test set, and verifying the fly recognition effect. According to the method, a fly face deep convolutional neural network is adopted, contour features of an image are roughly extracted by using a large number of convolution pooling, and then specific part features such as small eyes in fly compound eyes are extracted by using Inception-ResNet and Reduction networks. According to the network, the learning target and difficulty can be simplified, and richer feature vectors can be extracted while gradient loss is prevented.

The invention belongs to the technical field of information processing, and discloses a doctor-patient information display method and system, a storage medium, a computer program and a terminal. The method relates to user identification, user authorization, and information acquisition, release, display, browsing and query. The method comprises the following steps: a part of users are authorized toreceive electronic flags and business card information of other users according to user roles, and demand information encrypted by a system database is queried; a doctor-patient information display social platform identifies the user as a common user, a real-name user, a doctor user and an institution user after passing the system authentication, enhances platform information interaction of the users with the four identities, achieves user authority management, and encoding four types of information including demand information, doctor display information, institution display information andinstitution acquisition information according to user identification identities. Compared with other social platforms, the information is more real and accurate, and deeper understanding is provided for the user; massive real-time information retrieval is realized, data is effectively managed, and system statistical evaluation is also realized, so that user information interaction is facilitated.

The invention discloses a radar distance super-resolution method based on target sparsity and time domain dechirping. The radar distance super-resolution method comprises the following steps: 1) carrying out dechirping processing on a linear frequency modulation echo signal of a radar target in a time domain; 2) establishing a sparsity mathematical model of multi-target echo signals; and 3) constructing an optimization function about the target distance and solving the optimization function to obtain the target distance, and meanwhile, realizing radar distance super-resolution. Under the sameradar signal bandwidth, compared with conventional pulse compression processing, the method has a higher distance resolution capability, and distance super-resolution is realized; and the super-resolution in distance improves the performance of target resolution, imaging, classification and recognition, and target detection in a clutter.

The invention relates to a screening device for textiles. The screening device for the textiles comprises a bottom seat and a side seat; a gathering and dividing frame is mounted at the upper ends of the bottom seat and the side seat commonly; a linear sliding way is formed in the left side of the gathering and dividing frame; the left side of the middle part of the linear sliding way communicates with a V-shaped lateral sliding way; an annular blocking rod is arranged on the right side of the middle part of the linear sliding way; and the left end of the blocking rod is connected with a motor. The screening device for the textiles is convenient to use; and defective textiles and qualified textiles can be respectively fed into different containers on the way of feeding each textile into the containers, so that the effects of screening and classifying and improving the delivery speed of the textiles indirectly are achieved.

The invention provides a message sorting mechanism under a host overlay, which comprises the following steps of: 1, receiving a message under the overlay by a host, analyzing message data, identifying message content, generating a message header and a message body, and generating a session identifier through the message header; step 2, the controller accumulates the session identifier values generated by the message, and the net-agent service records the session identifiers after receiving the message; 3, further identifying and judging the session identifier; and 4, recording an index of the issued data, and executing a corresponding command on the processed configuration message. According to the method, verification is carried out at the sending end and the receiving end by using an algorithm, the transmitted message is further determined, accurate transmission of the message is better ensured, the message is issued through two paths, sequential execution of the message is ensured, and the efficiency is improved. According to the invention, even in a scene of large-scale deployment of virtual machine physical servers, the sequence of messages can still be ensured under the condition of passing through different data sources.

The invention discloses a method for analyzing flavor principal components of pickled cedrela sinensis subjected to different sterilization treatments, belongs to the technical field of cedrela sinensis processing treatment and cedrela sinensis flavor detection, and particularly relates to a method for analyzing flavor principal components of pickled cedrela sinensis subjected to sterilization treatments, enriching cedrela sinensis fragrance by using a headspace-solid phase microextraction (HS-SPME) technology and analyzing the flavor principal components of the pickled cedrela sinensis through gas chromatography-mass spectrometry (GC-MS). According to the method, the flavor substances of the pickled cedrela sinensis are rapidly identified in combination with SIMCA, 10 kinds of cedrela sinensis are pickled in different treatments, through PCA analysis in the SIMCA, the flavors of 10 kinds of cedrela sinensis products in different treatments can be effectively identified, and the superiority of different treatment methods is compared.

The invention provides an exposure machine which is convenient to maintain, comprises an exposure machine shell, a storage battery, an adjustable hose, a power switch, an auxiliary lighting lamp, a cover, a handle, a replaceable exposure lamp holder structure, a standby exposure lamp holder storage frame structure, an adjustable chassis structure, a fixed outer frame, an air guide pipe, a fan bracket and a micro axial fan, wherein the storage battery is glued on the upper part of the exposure machine shell; The adjustable hose is glued on the upper right side of the battery; The power switch is glued on the right side of the battery; The auxiliary lighting lamp is glued to the right end of the adjustable hose. An LED exposure lamp is embed on that upper side of the exposure lamp rack of the invention, which is favorable for dismantling the single exposure lamp rack when the LED illumination lamp on the single exposure lamp rack is damaged, and the device is convenient to maintain; Thestandby lamp holder is arranged in the storage tank, which is favorable for storing the standby lamp holder, so that the standby lamp holder can be supplied in time when the exposure lamp holder is damaged.