68 results about "Confidence region" patented technology

The invention provides a method for estimating parameters of a conductivity second-order resistance-capacitance coupled network, and belongs to the technical field of soft measurement of the solution conductivity, wherein the measurement of the solution conductivity is converted into parameter estimation of a second-order equivalent resistance-capacitance coupled network; and the method specifically comprises the following steps of: for a conductance cell second-order resistance-capacitance system model taking lead distributed capacitance and electrical double-layer capacitance influences into account, orderly employing two sine excitation signals of different frequencies to excite the resistance-capacitance network; on the basis that the amplitude ratio and phase difference of sine response signals in relation to the sine excitation signals thereof are estimated, respectively, establishing an over-determined system related to the resistance and capacitance parameters to be estimated according to the relationship deduced between the amplitude ratio and the phase difference and the solution resistance, the lead distributed capacitance and the electrical double-layer capacitance; and then obtaining the estimated values of the resistance and capacitance parameters by a subspace confidence region optimization algorithm based on the nonlinear least square principle. The method provided by the invention is high in rate of convergence, high in estimation accuracy and good in anti-interference capability, and suitable for industrial online real-time application.

A method and apparatus for detecting the presence of abnormal tissues of the present invention utilizing a magnetic resonance imaging system in communication with a computer. The apparatus is configured to define a confidence region using a probability based confidence interval calculation such as multivariate or bivariate analysis for at least two parameters in a normal tissue sample by making a magnetic resonance image, and then to evaluate a second sample by making a magnetic resonance image. Parameters detectable by magnetic resonance imaging 3 are selected from t1, t2, t2*, signal intensity, and diffusion.

The present invention relates to a confidence region algorithm-based multidimensional space contour error estimation method for solving the disadvantages that the conventional multidimensional space contour error compensation control precision is low and the convergence is greatly influenced by an initial value. The method of the present invention is characterized by solving a neighborhood of a current iteration point as an iteration domain at each time of iteration, obtaining a tentative iteration step length in the neighborhood, defining an evaluation function to decide the acceptation or rejection of the step length and the range of a next iteration confidence region, if the step length satisfies the requirements of the evaluation function, updating a current iteration state and keepingor expanding the confidence region, otherwise keeping an original iteration state and reducing the confidence region until the precision satisfies the requirements or stopping the iteration when theiteration number of times reaches an upper limit. Compared with a Newton's method, the method of the present invention guarantees the overall convergence, enables the derivative calculation to be reduced, and is used for the contour tracking and precise processing technology field.

The invention discloses a multi-dimensional field weed identification method based on generative adversarial learning. A semi-supervised learning strategy is adopted to train an overall network. The overall network is trained by inputting labeled data and unlabeled data. Tagged data is used as input of a discrimination network. training the network by using unlabeled data. A space probability graph generated by unsupervised data is used as calculation of semi-supervised loss. The method can effectively reflect the similarity between a prediction result and real data distribution, uses generative adversarial learning to find a confidence region of unlabeled data, obtains a final field weed recognition result through self-learning, uses a confidence graph to perform self-learning on a network, and uses a semi-supervised learning strategy to break through the limitation of insufficient labeled data. According to the invention, the segmentation network is used to replace a traditional network, input features of different sizes can be accepted, and the flexibility is high.

The invention discloses a processing method of a pavement crack image. The method mainly includes the following contents: 1, using an OTSU method to process the pavement crack image through an iteration process, and eliminating non-confidence regions in all iteration results in the pavement crack image; 2, executing a preprocessing algorithm which is of lesser iteration frequency and used on the pavement crack image; and 3, carrying out excessive iteration on the pavement crack image, processing a result, then using a processed result as a seed image, eliminating tiny regions in the seed image, utilizing an edge detection algorithm to set a ''growth fence'' in the pavement crack image, and finally completing image segmentation through regional seed growth. The pavement crack image processing method provided by the invention obviously improves an image preprocessing effect and an image segmentation effect, overcomes interference of noisy noises in the image, is higher in calculation efficiency, has higher robustness, is stable in performance, is applied to intelligent identification and statistics of pavement crack diseases, and improves identification precision and statistic efficiency of the pavement crack diseases.

The invention discloses a method for detecting and evaluating the compaction uniformity of an inorganic stabilized granular base course of a road. According to a detection point arrangement principle, the Evdi of a to-be-estimated section of the road is detected by adopting a PFWD (Portable Falling Weight Deflectometer); a detection result is subjected to statistic analysis, so as to obtain a deformation modulus average value and a variance; upper and lower confidence limits are calculated; the three-dimensional curved surface of Evd is cut; value lines of the Evd and the like are projected out; the area Sh of a qualified region and the area Sq of an overall region are calculated; a uniformity index t is calculated according to a definition; the compaction uniformity is estimated according to an evaluation standard; a judgment method of a compacted weak region of a roadbed is to define the compacted weak region of the roadbed by adopting an Evd lower confidence region of the to-be-detected section of the road; the quick aftercompaction can be carried out according to the position of the weak region.

The invention discloses a multi-view driving risk assessment method and a multi-view driving risk assessment device based on a network connection environment. The method comprises the following steps:step 1, dividing a region to be subjected to risk assessment into a plurality of sub-regions; 2, acquiring first environment sensing information of each sub-region through selected roadside sensing equipment, and judging the confidence coefficient of each sub-region; 3, judging whether the sum of the areas of the sub-regions corresponding to a low-confidence region in the region to be subjected to risk assessment is smaller than a preset ratio of the area of the region to be subjected to risk assessment or not; 4, fusing and splicing the first environmental perception information of the low-confidence region and the second environmental perception information of the same sub-region as the first environmental perception information to obtain an enhanced environmental perception result corresponding to the Alow of the low-confidence region; 5, generating a global environment perception result of the region to be subjected to risk assessment; and 6, evaluating the overall risk of the region to be subjected to risk assessment. According to the invention, the purposes of real-time perception and risk assessment without blind areas can be achieved by using a network connection environment.

The invention discloses an antenna optimization method capable of combining a moment method with a confidence region space mapping algorithm. The method comprises the following steps that: firstly, a rough model for an antenna space mapping algorithm is established, the construction of the rough model adopts a response surface approximation method based on a Kriging interpolation, the rough model is optimized, and the optimal design parameter of the rough model is determined; a fine model adopts a full wave analysis moment method, the response of the rough model is enabled to approach to the response of the fine model through parameter extraction, and a mapping relationship between a rough model parameter and a fine model parameter is established; and the optimal design parameter of the rough model and the inverse mapping of the established mapping relationship are utilized to obtain the predication parameter of the fine model, and if the predication parameter of the fine model does not meet design requirements, iterative update is carried out on the mapping relationship until the predication parameter of the fine model meets the design requirements. By use of the method, the parameter of the designed antenna is integrally optimized, and time is saved on the premise that accuracy is guaranteed.

The invention provides a dual-mode map matching method based on a three-evidence DS theory. The dual-mode map matching method comprises the following steps: selecting a matching mode, calculating a confidence region when a high-precision mode is executed, screening all candidate roads from the confidence region, calculating corresponding probability distribution function values for all the candidate roads by virtue of the three-evidence DS theory, arraying all the probability distribution function values in a descending order manner, and carrying out similarity verification from large to small; and carrying out dual verification on a history result by virtue of the three-evidence DS theory and the similarity when a selective verification mode is executed: firstly calculating the probability distribution function value of a front matching road by virtue of the three-evidence DS theory, and if the probability distribution function value is not smaller than a preset threshold value, carrying out selective verification. The dual-mode map matching method is superior to other map matching methods and has good robustness and accuracy, and the matching time can be shortened.

The invention discloses an image processing method add device, equipment and a storage medium, and relates to the field of image processing. The method comprises the following steps: processing an image through an image segmentation model to obtain a probability distribution diagram of a sky region in the image, namely the confidence coefficients of pixels of the sky region in the diagram; determining a high-confidence region of which the confidence coefficient is higher than a first confidencecoefficient threshold from the sky region; and when the pixel points in the high-confidence region donot meet the preset condition, replacing the sky region with a target sky material to obtain a target image. According to the method and device, the problem that the sky area can be manually replacedby at least seven steps in the related technology is solved, the purpose of automatically replacing the original sky area in the image by the target sky material is achieved, a user does not need torepair the image in the later period, the manual operation steps of the user are reduced, and the human-computer interaction efficiency is improved. Even if the user does not have a picture repairingfunction, the user can obtain sky pictures which are comparable with sky pictures shot by a professional digital single lens reflex.

Determining a suitable target for an entity (such as a product) in a process control environment, based on observed process control data. A preferred embodiment organizes data in a hierarchical structure designed for automating the target-setting process; derives target “yardsticks” for various components based on this data structure; employs techniques to estimate proportions using sample-size-based trimming in conjunction with bias-correction techniques (where appropriate); and derives targets based on combining yardsticks and confidence regions for parameters that characterize component quality

A ridge direction extraction unit, which analyses the shape of a ridge in a fingerprint image and extracts the ridge direction that indicates the slope of the ridge, includes: a low confidence region density value conversion module that reduces the density values in a low confidence region to generate a density conversion image; a synthesized image generating module for synthesizing a high confidence region and the low confidence region to generate a synthesized image; an auxiliary direction determining function for determining which ridge direction in the synthesized image is to be the auxiliary direction and deriving the degree of confidence in the auxiliary direction; and a high confidence region expanding module for calculating the degree of confidence in the direction, and correcting the ridge direction in the low confidence direction contiguous with the high confidence region so as to increase the degree of confidence in the ridge direction.

The invention relates to a method and device for enlarging a CT display view field. The method comprises the steps that an intercepted trajectory is extracted from a confidence region adjacent to an intercepted region; and the intercepted trajectory is externally plugged into a complete trajectory.