31results about How to "Less input data" patented technology

A method is provided for extracting outer space feature information from spatial geometric data. The method comprises: an input step of inputting spatial geometric data for a target region; a sampling step of determining a sample by selecting an arbitrary area for the spatial geometric data input in the input step using a preset selection method; a feature extraction step of acquiring feature information for a corresponding sampling plane using a convex hull method based on sampling information including sampling plane information of the spatial geometric data for a sampling plane selected in the sampling step. The sampling step and the feature extraction step are repeatedly performed in a preset manner.

The special yarn manufacturing apparatus of the present invention performs the speed change of the front roller based on a plurality of pattern data indicating changes in the thickness of the spun yarn, and manufactures a special yarn corresponding to the above pattern data including at least information on the distance between thick pitches. Wire. The CPU controls the first servomotor in accordance with an execution program for executing a repeating pattern of one cycle consisting of pattern data larger in number than the pattern data based on the pattern data stored in the working memory. , thereby spinning a spun yarn that performs a repeated pattern structure of one cycle.

The invention discloses a prediction method for the net photosynthetic rate of a population and aims to overcome the problems that it is difficult to establish a prediction regression equation for the net photosynthetic rate of a population, considerable related data are needed in establishment of a prediction model for the net photosynthetic rate of the population and prediction accuracy is not high. The method comprises the following steps: 1, acquiring proportioning data of spectral radiance of every wave band of visible light in a region, wherein a portable multispectral radiometer is installed at a height H higher than the canopy of a plant in the test region with an area of S, the proportioning data M_D of spectral radiance of every wave band of visible light in the test region in different periods is obtained, and [0,1] normalization processing is carried out; 2, acquiring data of the net photosynthetic rate of the population; 3, constructing a bionic kernel function; 4, establishing an SVM training set and an SVM prediction set; 5, carrying out tool selection and parameter optimization on a prediction model; and 6, predicting the net photosynthetic rate of the population, which comprises obtainment of the prediction model Model, obtainment of a predicted value Predict of the net photosynthetic rate of the population and determination of reliability of the prediction model Model.

The invention discloses a method for calculating the forced excitation hotspot temperature of the air-cooled hidden pole synchronous condenser rotor. By inputting the calculation model and the forced excitation parameters, the model geometry is firstly divided into discrete heat flow node topology connection graphs; each heat flow node is gradually determined. Parameters such as area, depth, thermal resistance, resistance, heating power, etc., list the energy conservation equations of each node with the node temperature as the unknown variable and form a system of equations, and obtain the forced hot spot temperature of the rotor winding of the air-cooled camera by solving the equations . Only need to input 24 parameters, it can automatically calculate the rotor forced excitation hot spot temperature, and automatically draw the heat flow node topology diagram, which can well meet the requirements of air-cooled hidden pole synchronous motor rotor forced excitation hot spot temperature calculation and curve drawing, to achieve convenience, The purpose of quickly and accurately obtaining the forced excitation hot spot temperature of the air-cooled rotor. It has the characteristics of less input data, simple operation, quick calculation and drawing, concise and intuitive topological connection diagram of heat flow nodes, etc.

The invention discloses a method for detecting a periodic defect of a calendered material surface in a calendering process. The method comprises the following steps: S1, collecting a defect source image of the calendered material surface to be detected; S2, extracting image features of the defect in the defect source image; S3, pre-judging the periodic defect of the defect source image; and S4, matching adjacent suspected periodic defects in the current longitudinal region by adopting an image relevance matching method one by one to determine the periodic defect. The method can be used for quickly and accurately detecting the periodic defect of the calendered material surface, so that the omission ratio and pseudo-flaw false drop rate of periodic defect detection under various environments can be improved. The method is applicable to any material.

The present invention relates to a method for estimating the queuing length of lanes at signalized intersections based on single-section electric police data. The time stamp information of the vehicle passing the stop line, including the departure time, direction number, lane number and vehicle model parameters; 2) According to the time stamp information and the time sequence of the vehicle leaving the headway, for single-lane layout conditions and multi-lane layout conditions, pass The characteristic function is defined to describe the dual characteristics of approaching distance and headway fluctuation, identify the change point of the departing headway sequence, and estimate the vehicle queuing length and missed detection rate in each lane within the period. Compared with the prior art, the present invention takes into account the phenomenon of false and missed detection in electric police equipment, and has the advantages of covering all scenarios, less input data, wide applicability, and the like.

The invention provides a method for plotting a water temperature-power output curve of a generator. The method is characterized by comprising the steps of: calculating a allowable largest excitation current value at a water temperature point with a certain number; setting a rated power output value as an initial power output value of the generator; calculating an excitation current value according to the power output value of the generator; increasing the power output of the generator if the calculated excitation current is larger than the allowable largest excitation current at the water temperature point with the certain number, otherwise, decreasing the power output of the generator; calculating a new excitation current value according to the new power output of the generator; obtaining the power output value of the generator corresponding to the water temperature point with the certain number when the error between the new excitation current and the allowable largest excitation current at the water temperature point is smaller than 1A; and plotting the water temperature values of all water temperature points and the corresponding power output values to obtain the water temperature-power output curve of the generator. The method provided by the invention can overcome the disadvantages in the prior art, and has the advantages of being few in input data, convenient to operate, and high in curve plotting speed.

The invention discloses a spherical panorama generating method based on an overlapping way, comprising the following steps: (1) establishing a target spherical spatial model on the basis of two half-spherical spaces formed by two fisheye images; (2) calculating the mapping relation between the points in the target spherical space and the points in the half-spherical spaces; (3) calculating pixel values corresponding to the points on a spherical surface O according to a fisheye lens projection model and the mapping relation between the points in the spherical space and the points in the half-spherical spaces; and (4) mapping the points in the spherical space to a spherical panorama according to the principle of spherical panorama generation. According to the invention, panoramic stitching is realized with less input data, stitching is carried out with only two fisheye images, the data source is simple, and the cost is low. In the processing process, the location relationship of overlapping areas is introduced effectively, and the relative positions of pixels are quantitatively calculated according to the relationship, which simplifies the calculation process and speeds up stitching. Finally, image fusion is performed by a gray maximum fusion way, and the quality of panoramic stitching is improved.

The invention relates to a method for estimating transit time of a car on a mixed-traffic site road section based on a reliability model. The method comprises the steps of: the step 1: taking each carpassing a bus stop road section as an observation sample, taking the moment when a car arrives at a bus stop starting point as a keyword, and matching buses, cars and non-motor vehicles within a period of time before the moment into the observation sample to construct car road section transit time sample data; the step 2: building a bus road section transit time reliability model at the bus stopin a mixed-traffic environment, estimating parameters of the reliability model, and finally obtaining a final mathematical expression of the car road section transit time reliability model; and the step 3: on the basis of the step 2, obtaining the real-time bus arrival flow, the car arrival flow and the non-motor vehicle arrival flow as the input variables of the car road section transit time reliability model to estimate or predict the currently arrived car road section transit time.

The invention belongs to the technical field of image information processing, and discloses a fire image recognition method based on deep learning, including: collecting smoke pictures and normal pictures in the early stage of fire as the training set and test set of convolutional neural network; The dark channel images constitute the final training set and test set; construct a convolutional neural model that can detect smoke; train the neural network to obtain a smoke detection model, and test and evaluate the performance of the model. Compared with the prior art, the present invention improves the correct rate of smoke detection in a single image by using dark channel images and deep learning methods, and at the same time increases the detection speed, and can be actually applied to fire detection in cities or forests.