370 results about "Defuzzification" patented technology

The invention discloses a fuzzy control based automotive intelligent cruise assisted driving system control method. The method includes the steps: firstly, acquiring speed VF of a front key target vehicle, relative distance Sr and real-time speed Vh of a cruiser; judging whether the front vehicle exists or not, and entering an intelligent following mode if yes; otherwise, entering a constant-speed cruise mode; for the intelligent following mode, taking relative distance deviation RD and relative speed deviation RV as input parameters, and taking an output linguistic variable as pedal increment u of an accelerator pedal or brake pedal at one movement; according to the fuzzy rule, determining a fuzzy value of an output variable u; adopting a centroid method to perform defuzzification computation to obtain an accurate controlled variable so as to control throttle percentage or brake pedal travel. By the control method, automatic regulation and trigger control of the throttle percentage and brake pedal travel can be realized at the same time, an integrated control function integrating ACC(active cruise control) and run-stop control can be realized, running safety of an automobile is improved, and driving fatigue of a driver is alleviated.

The invention provides a multi-robot path planning and coordination collision prevention method based on fuzzy logic. The method comprises the steps of confirming an input variable and an output variable of a fuzzy controller by setting up the fuzzy controller with robots as a system, using related languages to describe the input variable and the output variable of the fuzzy controller, confirming a qualitative reasoning principle according to the fuzzy control theory, setting up fuzzy control rules according to input signals and output signals of the fuzzy controller, selecting a subordinating degree function of each input language variable and a subordinating degree function of each output language variable, finally, conducting defuzzification, conducting voting on the basis of the maximum principle of the subordinating degree functions, conducting corresponding motions on the robots, and completing the task for planning a path of a single robot. On the basis of the method, the other robots are regarded as dynamic front barriers, and the aim of coordination collision prevention of the multiple robots can be achieved.

A fuzzy logic control arrangement is provided for an impedance match network of the type that is typically employed between a source of RF power at a given impedance, e.g., 50 ohms, and a non-linear load whose impedance can vary in magnitude and phase, e.g., an RF plasma. The fuzzy logic controller fuzzifies the phase and the magnitude error signals. The error signals are applied to a fuzzy logic interference function based on a number of fuzzy sets. The values of the error signals enjoy some degree of membership in one or more fuzzy sets. Fuzzy logic rules are applied to the phase and magnitude error signals. In a defuzzification stage, drive signal values are obtained for moving the tuning elements of the variable impedances. The drive signal values are weighted according to respective fuzzy inference functions for which the error signals enjoy membership. Then the weighted drive signal values are combined to produce output drive signals.

Aiming at the defects in the prior art, the invention discloses a control method of a speed regulator of a servo system of a flat knitting machine. A system in the prior art has low adaptability and low stability. According to a fuzzy proportional-integral (PI) control algorithm of the control method, the running speed of a transmission mechanism of the computerized flat knitting machine is used as a control object; the difference value between the practical reference speed and the feedback speed and the change rate of the difference value are served as input of a fuzzy controller; the input quantity is fuzzified through selecting an appropriate universe of discourse and an appropriate membership function; appropriate fuzzy rule tables are set by utilizing the practical tuning strategies of PI parameters; and after a Mamdani fuzzy reasoning algorithm and defuzzification processing are adopted, the variable quantities of parameter values of the PI controller are output, thereby realizing on-line correction of the PI parameters. By adopting the control method, the disadvantages of a traditional manual correction method for the PI parameters are overcome, and on-line real time correction of the PI parameters is realized, thereby improving the adaptability and the stability of the system.

The invention relates to a self-learning mechanism-base fast matching fuzzy reasoning method. The method includes the following steps that: a Gaussian membership degree function method is adopted to construct parameter fuzzification information; a fuzzy rule base is established; external parameters are fuzzificated, so that a fact item can be obtained; the fact item is matched with rules in the fuzzy rule base by adopting a rete algorithm, so that a fuzzy reasoning result can be obtained; the fuzzy reasoning result is subjected to defuzzification, so that a final reasoning result can be obtained; and a sample set is constructed according to the final reasoning result and an actual feedback result, and rule strength self-learning correction is carried out based on the sample set. According to the self-learning mechanism-base fast matching fuzzy reasoning method of the invention, the rete algorithm is adopted, so that the efficiency of fuzzy reasoning can be improved, and the fuzzy reasoning method can be applied to the engineering field with high real-time requirements.

The invention discloses a TDM MIMO based speed defuzzification method for a vehicle-mounted FMCW radar. The method comprises the following steps that a multi-input-multi-output (MIMO) antenna array isconstructed; transmitting antennas emit FMCW successively, and one MIMO period is realized when all the transmitting antennas complete emission; the last step is repeated to realize MIMOnum periods;2D FFT is carried out on an obtained MIMO difference frequency signal, and the distance and fuzzy speed of a target are calculated; and then Doppler phase compensation and Doppler fuzzy compensation are carried out on the target to obtain the real speed of the target. According to the method, a Doppler compensation factor is searched and calculated so that the third dimension FFT has no residual phase; and before the third dimension FFT, both Doppler phase compensation and Doppler fuzzy compensation are carried out, so that the problem of speed fuzziness of the FMCW radar due to TDM MIMO is solved, the speed measurement range is widened, the accuracy of angle measurement is ensured, there is no loss in the high-angle resolution of MIMO, and the method is more practical in the automobile radar field.

The invention discloses a method for controlling the frequency modulation of a micro-grid battery energy storage system based on fuzzy control, and belongs to the technical field of electric power system micro-grids. On the basis of the conventional proportion integration differentiation (PID) control, fuzzy control and an implementation mode thereof are introduced into the method, and the method comprises important components such as fuzzification, fuzzy rules, fuzzy reasoning, defuzzification, PID control and the like. According to the method, the micro-grid frequency deviation and the micro-grid frequency change rate are fuzzified into input of a fuzzy controller, PID control parameters of active power output control are output according to a fuzzy control rule, and an active power output reference value Pref of the battery energy storage system is finally output to control the active power output of the micro-grid battery energy storage system. Compared with the conventional PID control, the method has very strong adaptability to the switching of micro-grid parallel network / isolated network operational modes and to the nonlinearity and the time variability of electric network operational parameters, has a good dynamic response characteristic, and effectively improves the active power control accuracy of the battery energy storage system and the frequency stability of the micro-grid.

The invention discloses a vehicle speed tracking fuzzy control method of a vehicle robot driver, comprising the following steps: (1) determining an input variable and an output variable of a fuzzy controller of the vehicle robot driver, and carrying out scale transformation on an actual input variable to lead the actual input variable to be transformed in a domain range; (2) using the fuzzy controller of the vehicle robot driver to carry out fuzzification on the actual input variable after scale transformation to obtain a fuzzy value, and using the fuzzy value to establish an input and outputvariable membership function; (3) establishing a fuzzy control rule of the vehicle robot driver based on a learning method of a driving experience knowledge base and measuring data; (4) using a fuzzyreasoning and a centroid defuzzification method to determine a fuzzy control table of the vehicle robot driver based on the input and output variable membership function and the fuzzy control rule; and (5) querying the fuzzy control table of the vehicle robot driver to perform online real-time control on the vehicle robot driver. The invention features good instantaneity, strong anti-interferenceability and fine vehicle speed tracking control precision.

The invention discloses a vehicle light control method for an adaptive front lighting system. The method comprises the following steps of: determining the vehicle light turning angle of the adaptive front lighting system (AFS) by using an accelerator pedal, a brake pedal, a clutch pedal, a gear, a steering wheel turning angle and the height information of a front axle and a back axle; simulating the attitude and the speed of a vehicle according to the characteristics of an AFS kinetic model, and acquiring the turning angle of a vehicle light by combining different vehicle types and road conditions; creating a language control rule by combining the empirical ranges of the left, right and longitudinal turning angles of a front light, and performing fuzzy reasoning and constructing a fuzzy control rule table according to the language control rule; and acquiring a fuzzy output decision by calculating a fuzzy relationship, and acquiring the actual control quantity of a vehicle light turning angle through defuzzification so as to output a motor control value to a motor driving module.

The invention discloses a method for controlling a PID controller route based on fuzzy control. The method comprises the steps that on the basis of a motor PID controller, a route correcting strategy based on the fuzzy control idea is integrated, fuzzy PID controlled quantities of delta Kp, delta Ki and delta Kd are obtained by detecting deviation and the deviation regulating rate, after the fuzzy PID controlled quantities are received by the PID controller, the fuzzy PID controlled quantities are subjected to defuzzification to be adjusted to be three precise controlled quantities of Kp, Ki and Kd. Accordingly, a motor of self-tracking equipment is controlled, and the purpose of correcting the route is achieved.

The invention discloses a PID (proportion, integral, derivative)-type?fuzzy logic control method based on a weight rule table. The method comprises the following steps: 1) a PID?signal conversion unit carries out conversion on given signals and feedback signals; 2) a fuzzy set is built and the weight value of each fuzzy description variable is defined; 3) the attribution ratio of each fuzzy description variable is determined; 4) the attribution ratio of each fuzzy description variable is multiplied by the corresponding weight value of the fuzzy description variable and adding is carried out to obtain adding signals; 5) the adding signals are outputted to a control operation unit; 6) the control operation unit outputs the signals to an execution unit for execution, and the feedback signals are acquired to the PID?signal conversion unit; 7) step 1 to step 6 are repeated until the given signals are equal to the feedback signals. According to the PID (proportion, integral, derivative)-type?fuzzy logic control method based on the weight rule table, a traditional and complicated fuzzy rule table is replaced by the simple weight rule table, such that expert experience can be presented more simply and intuitively; the entire control method is optimized in no need of a defuzzification?unit; and minimal?overshoot and?oscillation exist in the control process.

The invention discloses an electric transmission line icing prediction model based on a neural network and a fuzzy logic algorithm. The model comprises the following steps: reading micro meteorological parameters to form a training sample, modifying the weight of a network, introducing a threshold, acquiring the fundamental component of the icing thickness, reading position information of a pole tower, establishing an altitude subordinating degree function and a large area moisture distance subordinating degree function, establishing an error correction subordinating degree function, forming a fuzzy rule bank so as to obtain correction coefficients through defuzzification, and combining the calculation result of the neural network and the fuzzy logic compensation result. The electric transmission line icing prediction model with geographical location information is high in prediction precision when being compared with a conventional global model and a single BP (Back Propagation) neural network, and has a good effect in practical application.

The invention discloses a photovoltaic grid-connected inverter MPPT (Maximum Power Point Tracking) system based on the admittance optimization algorithm. In the invention, the advantages of the conductance increment algorithm and the fuzzy control algorithm are combined, the admittance optimization algorithm is independently researched and developed, and the fuzzified transfer function and the defuzzification judgment are established. A photovoltaic array Pmax is calculated through a photovoltaic array P-U curve, the voltage level fluctuation can be fast followed, no complex algorithm is required, the maximum power of a solar battery can be fast tracked, and the requirement for MPPT control on the photovoltaic system in various occasions can be well satisfied. Test results verify that the algorithm has better effects such as compatibility with different systems, fastness and the like in comparison with the conductance increment algorithm. The results of a great deal of experiments and test show that the photovoltaic grid-connected inverter MPPT system based on the admittance optimization algorithm has excellent real-time performance and dynamic performance, and is stable in tracking without misjudgment, with tracking accuracy reaching up to 99%.

The invention aims to provide a fuzzy control method for an automatic gearbox, which comprehensively considers the will of a driver and automobile driving cycle. The fuzzy control method includes the steps of acquiring driver action information and vehicle information, performing fuzzy control, arbitrating gears and the like. The fuzzy control method has the advantages that the will of the driver and automobile driving cycle information are fuzzified, the automatic gearbox automatically recognizes the will of the driver and an automobile driving environment under the current driving cycle, multi-parameter dynamic fuzzy control is introduced, the fuzzy control method is applicable to the optimal process of power shift under complicated working conditions, and the process of power shift based on an expert system is directly obtained by means of defuzzification by using fuzzy rules formed by driving experience and an expert knowledge base.

The invention discloses a power transmission line icing detection method based on a deep neural network. Defuzzification, contrast improvement and the like in image processing are fully utilized. Themethod comprises the following steps: carrying out preprocessing operation on field images with uneven quality shot by a camera and taking the images as training and testing data of a deep neural network detection model, thus the model can cope with detection tasks under different field environments, the recognition accuracy of the model is superior to that of an existing power transmission line icing detection method, the detection speed is far higher than that of the prior art, and real-time monitoring of the line condition becomes a reality.

The invention discloses a voltage sag source identification method based on Mamdani fuzzy reasoning, belonging to technical field of voltage quality monitoring of a power system. The method comprises the steps of calculating the membership degree of three-phase voltage to balance, the membership degree of the voltage at the end of voltage sag to mutation degree and the membership degree to harmonic content; inputting the membership degrees to an Mamdani fuzzy reasoning system and substituting into fuzzy rules to calculate a corresponding fuzzy result; and using the center of gravity defuzzification for defuzzifying the fuzzy result to obtain the type of a sag source. The method takes into account of inherent differences of four different sag causes and is started from amplitude, mutation and harmonic, thereby avoiding the impacts of a plurality of uncertain features on the analysis accuracy of different sag sources; in addition, the method fuzzifies the input and the output quantities of the reasoning system, thereby avoiding the practice that the prior art needs to determine thresholds of all the feature values and further reducing the identification errors caused by the fixation of the thresholds.

The invention provides a fuzzy control-based special PID (Proportion Integration Differentiation) method for controlling temperature of a plastic extrusion device. By taking the traditional PID control for reference and introducing an algorithm and an implementation mode of fuzzy rules, the method comprises a few important parts such as parameter fuzzification, fuzzy rule reasoning, parameter defuzzification, PID controller and the like. The method comprises the following steps of: calculating the deviation between an actual temperature and a theoretical temperature and current deviation variation speed according to reference input and feedback signals, performing fuzzy reasoning in a fuzzy reasoning device by means of fuzzy experience, performing fuzzification to convert an input value into a membership degree by means of a membership function of an input fuzzy set, outputting a proportion coefficient, an integration coefficient and a differentiation coefficient of a PID controller by the fuzzy reasoning device, and performing PID control by using the coefficients as references of the current controller. By calling the fuzzy rules of the heating and water cooling processes of a plastic extruder, nonlinear effect produced by water vaporizing and cooling in the cooling process can be effectively controlled, the temperature control of the whole main machine meets the requirement of the extrusion temperature, the extrusion quantity of the plastics is stabilized, and the product quality is improved.

The invention relates to an image processing method and device, a computer readable storage medium and electronic equipment. The method comprises the following steps of: carrying out face recognitionon a to-be-processed image so as to obtain a face area in the to-be-processed image; if a first definition value of the face area is lower than a first threshold value, carrying out defuzzification processing on the face area; determining a beautifying grade according to a second definition value of the processed face area; and adjusting beautifying parameters according to the beautifying grade and beautifying the face area according to the adjusted beautifying parameters. According to the method, when the face area in the image is relatively fuzzy, defuzzification processing is firstly carried out on the face area in the image, the beautifying grade is determined according to the definition of the face area after the defuzzification processing, and then the face area is beautified, so that the condition of relatively bad image effect caused by beautifying the image when the face area is relatively fuzzy is avoided.

The invention discloses an image haze removal method on the basis of self-adaptive illumination calculation and relates to an image processing method. The image haze removal method comprises the following steps of: calculating a bright channel image roughly extracted from a haze image and then carrying out edge information preservation processing to obtain an optimized bright channel image; calculating a dark channel image roughly extracted from the haze image and then carrying out edge information preservation processing to obtain an optimized dark channel image; according to an atmospheric scattering physical model and the obtained bright channel image and dark channel image, deriving a self-adaptive atmospheric lighting and atmospheric transmission coefficient expression; and according to the haze image, atmospheric lighting and an atmospheric transmission coefficient in the atmospheric scattering physical model, carrying out haze removal recovery processing. The bright channel of the image is disclosed for the first time, calculation of the bright channel is carried out on the haze image and the atmospheric lighting expression is derived according to the atmospheric scattering physical model, so that the atmospheric lighting can be self-adaptively calculated. Compared with a conventional haze removal method, the image haze removal method disclosed by the invention has the advantage that a defuzzification effect of the recovered image can be improved.

The invention discloses an intelligent risk assessment method for electric power security risk assessment. The intelligent risk assessment method for electric power security risk assessment comprises the steps: step 1, determining an input risk factor set according to the operating type of a power supply enterprise; step 2, performing fuzzification processing of the risk factors in the input risk factor set, and determining the proportion of the risk factors in an assessment set; step 3, performing fuzzy inference of fuzzy rules between the risk factor and the risk assessment result, and calculating the relevance grade of each fuzzy rule; step 4, performing normalization processing of the relevance grade of each fuzzy rule for the risk factors; and step 5, according to the fuzzy rules and the relevance grade of each fuzzy rule, calculating output of each fuzzy rule and performing defuzzification processing of the output, thus obtaining the risk assessment result. The intelligent risk assessment method for electric power security risk assessment is provided with knowledge inductive learning ability, and can also automatically adjust parameters of a membership function, thus effectively improving the accuracy and the validity of the assessment result.

The present invention discloses a brushless DC motor fuzzy control system based on a genetic algorithm and a control method thereof. The control system comprises a drive system used for driving a brushless DC motor, a speed observer used for collecting the rotation speed of the brushless DC motor, a rotation speed reference model which provides the reference data of a motor rotation speed and compares the reference data of a motor rotation speed with the collected motor rotation speed so as to obtain a rotation speed error and an error change rate, a parameter fuzzification module which receives the rotation speed error and the error change rate, carries out quantification on the rotation speed error and the error change rate, and maps the rotation speed error and the error change rate to a fuzzy set discourse domain, a genetic algorithm optimization module which uses the genetic algorithm to carry out online optimization on a module control rule, adjusts related parameters in the fuzzy set discourse domain and makes a fuzzy decision, and a defuzzification module which maps the output amount of the fuzzy decision to a basic discourse domain. According to the system and the method, the genetic algorithm is used to carry out online adjustment of the parameter of the fuzzy controller, and the controller can have good static and dynamic performance in different operation environments.

Apparatus for controlling a land vehicle which is self-driving or partially self-driving, which apparatus comprises a coarse tuning assembly (1, 2, 3) and a fine tuning assembly (4), the coarse tuning assembly (1, 2, 3) being such that It comprises: a. a sensor interface (1) which measures kinematic parameters including speed and braking, b. fuzzy descriptions to model guidance, navigation and control of the vehicle, the fuzzy descriptions including: (i) driver behaviour and driving dynamics, (ii) uncertainties due to the environment including weather, road conditions and traffic, and (iii) input faults including mechanical and electrical parts, and c. an adaptive fuzzy logic controller (3) for nonlinear MIMO systems (2) with subsystems which comprise fuzzification, inference, and output processing, which comprise both type reduction and defuzzification, and which provide stability of a resulting closed-loop system, the adaptive fuzzy logic controller (3) including: (i) inference engine identifying relationships using a rule base and outputs as 'fuzzy sets' to a type reducer, and (ii) output control demands including torque actuators to the fuzzyfier 'fuzzyfiying' the signal, and the fine tuning assembly (4) being such that it comprises: a. inputs from the coarse tuning assembly (1, 2, 3), b. precognition horizons determining how many future samples the objective function considers for minimization and the length of the control sequence computed, c. a linearized MIMO regression model extracted from the adaptive fuzzy logic controller (3) at each time step providing the 'fine' tuning parameters, and d. a non-linear dynamic linearized regression controller (4a) providing: (I) a crisp output signal feeding into APACC synthesis (4b) computing the optimal future vehicle guidance, navigation and control sequence, and (II) reduced set output and APACC synthesis (4b) feeding into the APACC linear logic system.

A steering control method including the steps of obtaining a heading error, obtaining a velocity value, obtaining a distance error, applying the heading error and defuzzrfying an output from a steering rule base. The velocity value and the distance error are applied along with the heading error to fuzzy logic membership functions to produce an output that is applied to a steering rule base. An output from the steering rule base Is defuzzified to produce a steering signal

The invention discloses a PID (proportion integration differentiation) parameter tuning method and tuning system. The PID parameter tuning method includes the steps: acquiring an initial PID parameter; acquiring an input variable to reflect control effect of a PID controller on a controlled object according to the initial PID parameter; correcting the initial PID parameter according to the control effect. The particular correcting process includes the steps: querying a preset fuzzy rule in a fuzzy set theory domain by the aid of the input variable to obtain a PID correcting parameter in the fuzzy set theory domain; performing defuzzification for the PID correcting parameter in the fuzzy set theory domain to obtain a PID correcting parameter; correcting the initial PID parameter by the aid of the PID correcting parameter to obtain a PID tuning parameter. The initial PID parameter is corrected according to the control effect of the initial PID parameter, so that the controlled object is more accurately controlled by the PID controller.

The invention provides a multi-loop fuzzy coupling temperature and humidity controller suitable for a constant temperature and humidity experiment box. Compared with the traditional temperature and humidity controller for the constant temperature and humidity experiment box, the controller has a distributed structure mode; independent acquisition circuits are adopted in temperature and humidity control respectively; coupled calculation of temperature and humidity parameters is performed in a bus communication-based mode; algorithms and realization modes of fuzzy rules comprising parameter fuzzification, fuzzy rule reasoning, parameter defuzzification and the like are introduced; membership degrees of transition temperature and humidity input values are calculated according to temperature and humidity reference input and feedback signals and a membership function of a temperature and humidity input fuzzy set; and meanwhile, a compounding proportion integration differentiation (PID) control function of temperature and humidity is called on the basis of a coupling model of temperature and humidity, and the proportion coefficient, the integration coefficient and the differentiation coefficient of a temperature and humidity PID controller are calculated and used as references of the current controller for controlling the temperature and the humidity. By practical application of the multi-loop fuzzy coupling temperature and humidity controller for the constant temperature and humidity experiment box, the control effect on the temperature and the humidity of the constant temperature and humidity experiment box can be effectively realized, the constant temperature and humidity requirements of the experiment box are met, the stabilities of the temperature and the humidity are higher than those of the traditional temperature and humidity controller, and the temperature index and the quality of the constant temperature and humidity experiment box are improved.

The invention discloses a fuzzy neural network model based on a multivariable phase-space reconstruction theory and an intelligent prediction method for deep excavation deformation by using the fuzzy neural network model. The fuzzy neural network model consists of four functional modules including a fuzzification interface, a fuzzy rule knowledge base, a fuzzy inference machine and a defuzzification device. The prediction model and the prediction method have high accuracy, and can effectively avoid major losses of countries and people's lives and properties.

The invention discloses a distance measuring defuzzification method for a pseudo-random code phase modulation continuous-wave radar, which is used for measuring a target distance in the field of continuous-wave radars. In the method, phase modulation is performed on carrier waves alternatively on circles by using a stagger pseudo-random code, and distance fuzzy resolution is performed by using a target distance remainder of two adjacent circles obtained by measuring, so that the real distance of a target can be determined. The method can replace the conventional pseudo-random code real-time alternate measuring method, and the detection and processing performance of the radar can be enhanced.

The invention belongs to the technical field of radars, and discloses an MTD radar defuzzification and speed measurement method. The method comprises the following steps that in a CPI, pulse group signals with different pulse repetition periods are sequentially transmitted by an MTD radar, echo signals corresponding to the pulse group signals are received, an echo data matrix corresponding to each pulse group signal is obtained, moving target detection is carried out on the echo data matrix compressed by all pulses in a DFT channel and an FIR channel respectively, a first Doppler frequency and a second Doppler frequency are obtained respectively, whether echo signals of a target are in a clutter area or a noise area is judged, the Doppler frequencies obtained by one of the channels are selected as the target Doppler frequency according to a judgment result, and the target speed is further determined. The method can improve the robustness of the system and the resource utilization rateof the radar, so that the signal-to-noise ratio of the target signal is improved, and the detection probability of the target signal and the success rate of speed ambiguity resolution are improved.

The invention discloses an intelligent least-square system and method for optimizing incinerator temperature of a pesticide waste liquid incinerator. According to the method, a least-square support vector machine is adopted as a local equation of a fuzzy system, defuzzification output is carried out on the output of the least-square support vector machine, and accurate control over the incinerator temperature is achieved. In the system and the method, training samples are processed through a standardization processing module and used as the input of a fuzzy system module; an incinerator temperature predicted value obtained from the fuzzy system module and an operating variable value optimizing the incinerator temperature are connected with a result display module, and the result display module is used for transmitting results to a DCS; a model updating module is used for collecting the signals of a field intelligent instrument according to a set sampling time interval. According to the system and the method, the fact that the incinerator temperature is calculated in real time and controlled accurately is achieved, and overshooting of the incinerator temperature is avoided.

The invention discloses an edge-based multi-direction weighting TV and self-similarity constraint image defuzzification method. A conventional TV model is improved based on edge detection, and transform domain self-similarity of an image is combined. Firstly, pixels in a center pixel neighborhood is divided into same-side pixel pairs and different-side pixel pairs by edge detection, different weights are adopted for different types of the pixels to obtain a multi-direction weighting TV algorithm, and the detailed information of the image is reserved to the greatest extent in defuzzification; secondly, a transform domain self-similarity regularization term is fused into the TV model, so that the limitation that the conventional non-local regularization depends on a non-local weight matrix is overcome; therefore, the texture and structural self-similarity of the image can be described more accurately; and the defuzzification visual effect of the image is further improved. By adoption of the defuzzification method, the visual effect is improved by defuzzification while the edge information and important details of the image are well reserved; and in addition, the objective indicators, such as a peak signal to noise ratio and the like, of the image are improved.