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

The invention provides a method for detecting smoky vehicles based on a single-frame image, which includes: converting a single-frame color image into a grayscale image, and dividing it into several small squares on average; Binary pattern features or discriminative non-redundant local binary pattern features; for each small square region, extract gradient position direction histogram features; for each small square region, extract local fractal features; normalize different types of Features, and concatenated to form the final feature vector, use the SVM classifier to classify the proposed feature vector, identify the small square with smoke, and make a judgment on whether there is a black smoke car in the current frame. The present invention introduces various features, so that the information expression is both sufficient and discriminative, more conducive to classification and recognition, and has stronger robustness and discriminative properties. The recognition rate is further improved and the false alarm rate is reduced.

The invention discloses a method for classifying SAR targets based on decision fusion of CNN and SVM. The idea is: determine a SAR image, the SAR image includes λ-type targets, and divide the λ-type targets to obtain a training set and a test set; Then standardize and subset the test set to obtain the first test subset testx1, the second test subset testx2 and the third test subset testx3; then obtain the standardized feature matrix traindata of trainx, the standardized feature matrix testdata1 of testx1, testx2 After the standardized feature matrix testdata2 and testx3 of the standardized feature matrix testdata3, the parameter matrix W and the offset vector bb are respectively obtained; the first column vector C1, the second column vector C2 and the third column vector C3 are determined; and the final predicted category column vector is obtained ; calculate v 2 class is recognized as v 1 The class probabilities are postscripted as the result of SAR object classification based on CNN and SVM decision fusion.

The invention provides a traceable safe unloading system for liquid hazardous wastes. According to the traceable safe unloading system for the liquid hazardous wastes, inlets and outlets are arrangedwithin a certain area; a material temporary storage area is arranged in the area and the material temporary storage area is divided into a plurality of small areas; the small areas are marked by different colors and each small area is divided into a qualified material area and a unqualified material area; material barrels corresponding to the colors of the small areas are placed in the corresponding small areas and radio-frequency identification codes are arranged on the material barrels; conveying belts are correspondingly arranged at the two sides of the material temporary storage area; a plurality of material unloading areas, barrel washing areas, drying areas and leakage detecting areas are sequentially arranged on the conveying belts and empty barrel storage areas are arranged at thetails of the conveying belts. According to the system, hazardous waste containers are unified and the radio-frequency identification codes are arranged, therefore the functions of being traceable without the need for manual recording and convenient for taking trace can be realized; the full-automatic continuous operation of material unloading and cleaning of the material barrels can be realized, so that the working efficiency is improved and the safety is guaranteed.

The invention relates to a magnetic tile surface defect detection system and method, and relates to a magnetic tile surface defect detection collar. The system includes: an image acquisition device and a computer system; the image acquisition device is used to collect the surface image of the magnetic tile, and the computer system judges the suspicious area on the surface image of the magnetic tile and divides the suspicious area to obtain the suspicious area image, and extract the geometric features, grayscale features and texture features of the suspicious area image, and the computer system inputs the geometric features, the grayscale features and the texture features into a pre-trained classifier to obtain The detection results of magnetic tile surface defects, through the classification recognition algorithm, realize high-precision classification and identification of magnetic tile defects, and realize automatic detection of magnetic tile defects, reduce labor costs, improve efficiency and recognition accuracy.

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 relates to a twitter identification method based on self-adaption energy detection under a complex environment. The method is characterized by comprising the following steps of firstly, dividing twitter into three frequency bands according to the frequency distribution condition of the twitter; secondly, detecting twitter acoustical signals containing noise through a self-adaption energy detection method based on noise power spectrum estimation and the foreground sound existence probability, and screening out the useful twitter signals; thirdly, extracting Mel-scale WPSCC features from the useful twitter signals according to distribution of the Mel scales; fourthly, conducting modeling, classification and identification on the extracted Mel-scale WPSCC features and extracted Mel-scale MFCC features through an SVM classifier. The WPSCC features extracted through the method have a better anti-noise function and are better in identification performance after undergoing self-adaption energy detection, the method is more suitable for twitter identification under the complex environment and has a good classification and identification effect on the twitter sound containing noise under the complex environment.

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 relates to the technical field of data processing, and provides a method and device for obtaining micro-motion characteristics. The method for obtaining the micro-motion characteristicscomprises the steps that radar echo signals of a warhead target are acquired; the time-frequency analysis of the radar echo signals is carried out to obtain a signal energy distribution function; according to the signal energy distribution function and the radar echo signals, at least one micro-motion characteristic of the warhead target is extracted; and a SVM classifier is used for filtering theextracted micro-motion characteristics, and an optimal feature subset including at least one of the micro-motion characteristics is obtained. The device for obtaining the micro-motion characteristicsincludes an acquisition unit, a time-frequency analysis unit, a characteristic extraction unit and a characteristic filtering unit. According to the method and device for obtaining the micro-motion characteristics, the micro-motion characteristics can be obtained from the radar echo signals.

The invention discloses an ISAR target contour extraction method based on the CLEAN algorithm scattering center extraction. First step 1: using the two-dimensional CLEAN algorithm to extract the scattering center position information of the ISAR target, and reconstructing the ISAR image according to the extracted N strong scattering points Get the final high-resolution image, use α-shape to get the scattering center position on the edge of the target; connect these scattering centers into a closed curve through the rubber band method proposed in this paper, and smooth the result to obtain a smooth target outline Wire. The beneficial effect of the present invention is to apply the two-dimensional CLEAN high-resolution technology to the ISAR image contour extraction, connect the scattering centers at the edge of the target with a closed smooth curve to obtain the ISAR target contour extraction through the rubber band algorithm proposed herein, It is convenient for subsequent ISAR target classification and recognition.