30results about How to "Fully express" patented technology

The invention discloses a polarization SAR image classification method based on multi-feature fusion. The present invention firstly extracts the polarization feature vector of the polarimetric SAR image to be classified to obtain a high-dimensional polarization feature set; extracts the morphological section feature vector of the image SPAN processing result to obtain a high-dimensional morphological feature set; After the features are subjected to the dimensionality reduction processing of the locality-preserving discrimination analysis, the image pixels with known category labels are selected to form the training sample set, and then the remaining pixels of the entire image are selected as the classification sample set; the two types of low-dimensional features are respectively passed through the The SVM of the maximum posterior probability obtains the category label and the corresponding posterior probability of the pixel points in each case; the summation criterion or the adaptive weighted summation criterion is used to combine the posterior probability vectors of each pixel in the two situations According to the principle of maximum a posteriori probability, the final classification result of high-resolution polarimetric SAR images is obtained. The implementation of the invention helps to improve the classification accuracy and efficiency of resolving polarimetric SAR images.

The invention discloses a visual SLAM algorithm that adopts CNNs feature detection in the whole cycle. First, at the front end, the original image data is pre-trained with an unsupervised model, and then the pre-trained data is used to combine the joint representation of motion and depth with the CNN network architecture. Local velocities and orientation changes are correlated to perform visual odometry; finally, path prediction is performed. The present invention also uses the OverFeat neural network model to perform the loopback detection link, which is used to eliminate the accumulated error brought by the front end and construct a visual slam architecture based on deep learning. At the same time, time and space continuity filters are constructed to verify the matching results, improve the matching accuracy and eliminate false matching. The present invention has great advantages and potentials in improving the accuracy of the visual odometer and the correct rate of closed-loop detection.

The invention relates to a label semantic learning method based on network structure and semantic correlation measurement, including initializing the label network based on user behavior facts to obtain a fact label network G; constructing a regulated label network G according to the fact label network G R ; in the label network G R Applying the improved random walk strategy to build a label network G based on the random walk strategy C ; Build a tag network G based on tag-related text information T ; for label network G C , label network G T Perform normalization processing, and learn label semantic vector representation through random walk strategy and word vector learning method. The invention has a reasonable design, not only makes full use of the network topology structure, but also takes into account the relevant text information contained in the nodes, and can learn from the topology structure and text expression to an easy-to-operate, high-confidence, full-expression, and low-noise system within a relatively short period of time. The tag semantic vector can be widely used in tag network learning and tag semantic learning of tagged text collections.

The invention relates to an optical information encryption method and encryption system. The system at least comprises an electrical encryption module and an optical modulation module, the electricalencryption module encrypts an electrical signal carrying data information into a first encrypted pulse electrical signal based on at least one encryption function, the optical modulation module modulates the first encrypted pulse electrical signal into an optical signal comprising at least two light emitting stages within one light emitting period based on the at least one optical modulation scheme, the light in the at least one light emitting stage independently changes according to a time-dependent characteristic curve, the characteristic curve corresponds to the first encrypted pulse electrical signal, the light in the at least one light emitting stage independently changes according to a characteristic curve associated with the encryption function of the first encrypted pulse electrical signal, and / or, the light in the at least one light emitting stage changes with the time at a characteristic curve close to zero konder. According to the optical information encryption method and encryption system, the light emitting period is divided into at least two light emitting stages, and the light emitting parameter changes in the light emitting stages are independently regulated and controlled to form encrypted light signals, so that the security level is high and is difficult to crack.

The invention relates to a semantic enhanced transport vehicle acoustic information fusion method, which comprises the following steps of: acquiring sound emitted by a running vehicle by utilizing a microphone array; extracting two groups of characteristics, namely harmonic characteristics X1 and key spectrum characteristics X2 from acoustic signals of the vehicle; marking available semantic attributes on the acoustic signals of the vehicle; and finally, arbitrating binary characteristic decision fusion by utilizing a semantic concept, and adopting a characteristic dimension distribution scheme when the signals are sent from a wheeled vehicle and adopting another characteristic dimension distribution scheme when the signals are sent from a tracked vehicle according to different semantic attributes. Dimensions of different characteristics are adjusted by analyzing the semantic attributes in the acoustic signals, different weighting effects are achieved, and the difference of characteristic importance is reflected, so that an effect of information fusion is improved. The accuracy of an acoustic transducer fusion system in the aspects of target identification or tracking can be improved by utilizing semantic analysis.

Disclosed are an animation synthesis method and device. According to the method, a terminal can receive text information inputted by a user and identify text keywords from the text information; then, the terminal can determine animations corresponding to the text keywords from a preset animation library, and synthesize the animations according to the sequence of the keywords in the text information, to obtain a fused animation. Compared with text information, an animation can more fully and vividly express the meaning of information. Therefore, compared with the approach in the prior art of only presenting information in the form of a text or voice, an animation converted from text information can more fully and vividly express the real meaning of the information, brining pleasure and convenience to a user during an information reading process.

The invention provides a pumping station operation information visualization method for energy efficiency evaluation, including determining the reference flow value and reference head value of the pumping station, obtaining the flow value, head value and efficiency value of the pumping station during operation, and sorting out the pumping station A record table composed of flow normalized value, head normalized value and efficiency value during the operation process, determine several operating characteristic points and their corresponding frequency values ​​and average efficiency values, restore the flow value and head value of the operating characteristic points and Perform steps such as visual drawing of the combination information of each operating feature point. Based on the statistical analysis of data, the present invention expresses the complex historical operation data of the pumping station into a number of representative operating characteristic points, and each operating characteristic point corresponds to statistical information such as frequency value and average efficiency value, and draws different positions, The solid circles of radius and color are used to distinguish different operating feature points, so as to complete the visualization of pumping station operating information, which has the advantages of being scientific, reasonable, intuitive and concise.

The invention provides a traffic flow prediction method based on a balance index smoothing method and a stack self encoder. The method comprises the following steps of (1) generating a training sequence and a test sequence; (2) performing preprocessing on the training sequence and the testing sequence; (3) obtaining a balance index value sequence by using the balance index smoothing method; (4) calculating the normalization value of each value in the balance index value sequence of a complete training traffic flow sequence by using a minimum maximum normalization formula, and obtaining the normalization value sequence of the complete training and test traffic flow sequence; (5) building a stack self encoder neural network; (6) training the stack self encoder neural network; and (7) testingthe normalization value sequence of the complete test traffic flow data sequence. By using the method provided by the invention, the complicated nonlinear characteristics of the traffic flow data arealso considered on the basis of considering the high-random performance of the traffic flow data; and the prediction precision of the traffic flow is improved.