41 results about "Cfar detector" patented technology

The invention relates to a self-adaptive constant false alarm rate target detection method. The method includes the following steps: 1), transmitting data received by radar into a matching filter; 2), transmitting signals output by the matching filter into a square-law detector for processing; 3), transmitting signals output by the square-law detector into a CFAR (constant false alarm rate) detector for processing, and acquiring an estimation value Z of cluster power level generated by a reference unit according to a corresponding CFAR algorithm through sampling; 4), outputting a final judgment by the CFAR detector according to the estimation value Z acquired in step 3), namely detecting whether a target exists in the unit or not. In the method, according to statistical average value and variance of sampling values in a reference sliding window, the sampling value with the variance larger than a certain numerical value is deleted, an average value of remaining valid sampling values is used to replace the sampling value, and an average value of the sampling values is recalculated.

The invention discloses a CFAR detector suitable for complex inhomogeneous clutters, belongs to the field of radar signal processing, relates to a radar self-adaptive detection technology and in particular relates to a self-adaptive constant false alarm rate (CFAR) detection method. A radar radiation region is divided into even small regions by use of a morphological classification algorithm, each small region serves as a reference window, the reference windows of the same terrain are fused, and the terrains of the various reference windows after fusion are numbered to obtain the terrain classes of the whole radar radiation region. According to the CFAR detector suitable for complex inhomogeneous clutters, cells to be detected and resolution cells in the reference window are independent identically distributed even in an inhomogeneous background, and the performance of the detector is not affected by the inhomogeneous background. As a result, the CFAR detector suitable for complex inhomogeneous clutters is capable of solving the problem of acute drop of the performance of the CFAR detector when the radar scans a complex region in practical application, and has the outstanding advantages of high accuracy and good practicability.

The invention relates to a radar target and shadow segmentation method based on wavelets and constant false alarm rates. The method includes the following eleven steps: inputting a synthetic aperture radar image; selecting a wavelet function; conducting wavelet multi-dimensioned decomposition; selecting a high frequency sub-image after wavelet decomposition; inputting a CFAR detector I; segmenting the selected sub-image; performing wavelet inverse transformation; removing a mean value; inputting a CFAR detector II; inputting a target that finally needs to be segmented; and segmenting the image again. The advantages of the method are that the influence of spot noise on an SAR image is reduced; the defect that a contrast ratio between a target and a background has to be high in terms of a single CFAR detection method is overcome, and a target zone and a shadow zone can be detected at the same time; weak scattering targets in the SAR image can be effectively segmented, and the method exhibits high universality. An experiment shows that the mentioned SAR image segmentation method based on CFARs and wavelet transformation is a feasible and effective segmentation method, and has great application prospects.

The invention discloses a radar signal detection method based on a convolutional neural network. According to the method, a distance-Doppler frequency spectrum obtained after pulse Doppler processingis used as an image sample, different characteristics of a time-frequency spectrum with or without a target are taken as identification characteristics, radar signal detection is realized by utilizingthe classification capability of a convolutional neural network, existence of a target signal is judged, the convolutional neural network is applied to radar signal processing, and signal detection is realized from the perspective of image processing. Compared with a traditional CFAR detector, the method of the inventiom has the advantages that the detection probability is improved, the influencecaused by the correlation of background clutters when the CFAR detector is used is avoided, and the limitation that the fixed CFAR detector is suitable for a specific environment is improved.

The invention relates to the technical field of radar target detection, and in particular, relates to an FPGA-based two-dimensional ordered statistical constant false alarm detector implementation method. According to the method disclosed by the invention, detection is carried out in each clock period from two-dimensional power matrix input, so that the repeated reading time delay of the data cache is reduced, and the detection speed can be effectively improved. Aiming at the condition of large calculated amount in a traditional OS-CFAR (Ordered Statistical Constant False Alarm Rate) sorting process, the FPGA implementation method is completed by adopting parallel comparison, so that the sorting complexity is reduced. The reusability of a module is fully considered, two-dimensional slidingwindow structure setting is completed through a shifting register set, two-dimensional OS-CFAR detection is achieved, and BRAM memory resources and DSP resources of the FPGA can be saved.

The invention discloses a method for simultaneously tracking and jamming based on RF stealth. The method comprises the steps of: firstly, analyzing a relationship between a radar detection distance and an interception distance of an enemy passive system, and acquiring a transmission power range required by a radar for detecting a target on the premise of ensuring that a radar signal is not intercepted; secondly, adopting a Kalman filtering and covariance control method, increasing a radar interval to reduce radiant energy on the premise of ensuring radar tracking precision; thirdly, analyzinga relationship between a detection probability of a constant false alarm rate (CFAR) detector and power subjected to blanket jamming, so as to obtain a jamming power boundary condition of effective jamming for the enemy passive system; and analyzing a relationship between detection performance of the CFAR detector and a number of jamming pulses, reducing the number of jamming pulses on the premiseof ensuring the jamming effect, and making the jamming pulses do not overlap with radar echoes. The method provided by the invention reduces the radiation energy under the condition of satisfying theradar detection performance and the jamming shielding performance, and improves the RF stealth performance of the airborne radar.

The invention discloses a CFAR detector based on KL divergence unit screening, and relates to the technical field of radar target detection. In allusion to the problems that the existing detector needs the priori information of the number of interference targets under a multi-target background, and the large-number deletion number in TM-CFAR needs to be manually set and cannot be effectively extracted from the environment, the invention designs a novel detector by combining KL divergence and an Otsu method. Abnormal values with relatively high amplitudes can be effectively eliminated on the basis of sequencing the detection units, the large-value deletion number in the TM-CFAR is adaptively and effectively determined according to the change of the environment, the number of the anti-impacttargets is adaptively adjusted according to the condition of a reference unit, prior information of the number of the interference targets is not needed, and the method has good detection performancein a multi-target environment.

The invention discloses a RATM-CFAR detector and detection method based on distance and angle joint estimation, and the method comprises the steps: collecting radar echo data, carrying out the 2D-FFT of the data to obtain an RDM distance Doppler matrix, employing a distance-angle estimation result for the priori estimation of a CFAR algorithm in a process of carrying out the CFAR of the RDM matrix, converting a frequency domain into a spatial domain, and carrying out the processing; eliminating interference targets near the to-be-detected unit and prerequisite targets far away from the to-be-detected unit, and then performing CFAR algorithm selection based on background clutter estimation on the remaining units to obtain a final detection result. According to the method, the distance-angle estimation result is used for priori estimation of the CFAR algorithm for the first time, conversion from the frequency domain to the spatial domain is achieved, the method has the advantages of being simple, capable of avoiding missing detection of small targets, high in detection precision, high in environmental adaptability and high in anti-interference performance, the detection performance of the radar in a complex environment and a multi-target environment is improved, and a new method is provided for the CFAR algorithm.

The invention discloses a radar-based sleep characteristic event detection method, device and a storage medium. The method comprises the steps as follows: S10, transmitting waveform to a sleeper in adetection range of radar and receiving a target echo signal by radar, establishing an echo signal model; and S20, preprocessing the received echo signal to obtain a restrained echo signal. The radar-based sleep characteristic event detection method fits distribution of power burst curves of sleep events and sleep characteristic events by matlab by analyzing power burst curve distribution of the sleep events and sleep characteristic events, judges whether distribution of P(Z|H1) and P(Z|H10) obeys lognormal normal distribution, and then detects numerical values of power burst curves by an adaptive logarithm double-threshold constant false-alarm detector; and the technical scheme can determine that power burst curve characteristics conform to normal distribution after taking logarithm valuesand accordingly, designs the adaptive logarithm double-threshold constant false-alarm detector to detect sleep characteristic events.

The invention relates to a target detection method of a matrix CFAR detector based on tKL divergence. The method comprises the following steps of: calculating a covariance matrix according to echoes received by each distance gate; calculating a right median matrix of the covariance matrix corresponding to a reference distance gate according to the definition of the tKL divergence; calculating a covariance matrix corresponding to a target to obtain the tKL divergence of a target covariance matrix and the right median matrix; determining a detection threshold value; and comparing the tKL divergence of the target covariance matrix and the right median matrix with the detection threshold value, and if the target tKL divergence is greater than the detection threshold value, determining that thetarget exists; and otherwise, determining that the target does not exist. By adoption of the target detection method of the matrix CFAR detector based on the tKL divergence provided by the technicalscheme, the problems of high difficulty and low accuracy of the existing target detection method for target detection under complex non-uniform clutter and short pulse sequence conditions can be effectively solved.

The invention provides a sea surface weak target detection method based on an AR spectrum singular intensity function, and the method comprises the steps: estimating an AR spectrum of a sea clutter through employing a time domain echo signal of the sea clutter, judging the fractal characteristics of the AR spectrum of the sea clutter, calculating the value interval width of the singular intensityfunction of the AR spectrum of the sea clutter, and carrying out the design of a constant false alarm detector. According to the method, the non-Gaussian, non-stationary and non-linear characteristicsof the sea clutter are utilized, and the defect that the detection performance of a traditional target detection method is reduced due to sea clutter model mismatch is overcome by analyzing the fractal characteristics of the AR spectrum of the sea clutter; AR power spectrum analysis is adopted, the defects of traditional time domain and frequency domain fractal analysis are overcome, the influence of time correlation on fractal characteristic analysis is fully considered, and the frequency resolution of the AR power spectrum is greatly improved; and the detection performance and stability ofthe sea surface weak target under the low signal-to-clutter ratio background are improved.

The invention discloses an improved two-dimensional change index constant false alarm target detection method. The method comprises the following steps: selecting a reference unit on a two-dimensionalR-D spectrum, and building a left reference window, a right reference window, an upper reference window and a lower reference window; describing clutter states and summing according to reference unitvalues in the reference window; selecting corresponding logic according to the background estimation condition; and calculating a detection threshold according to the weight to realize constant falsealarm processing. According to the method, an existing single-dimensional VI-CFAR detection strategy is expanded to a two-dimensional R-D spectrum for 2D-VI-CFAR detection, so that the background estimation effectiveness is improved, and the method has higher adaptability to various complex clutter environments. According to the invention, the respective excellent detection performances of the three mean value type constant false alarm detectors in a uniform clutter, a multi-target environment and a clutter edge environment are integrated, and the background reference unit is selected by using the two-dimensional spectrum information to estimate the background clutter power so as to select a more appropriate detection strategy.

The invention discloses an intelligent clutter partitioning method based on false alarm preprocessing time, and the method comprises the steps: step 1, carrying out average constant false alarm detection of a leading edge reference unit and the average CFAR detection of a trailing edge reference unit (Back of Off, BO) of each frame of data (kth frame) outputted by a detector at the same time; step2, fusing the detection results of the FO-CFAR detector and the BO-CFAR detector; step 3, if frame data is the first frame data, initializing a time storage matrix, otherwise, deleting the data in the lowest row of the storage matrix, integrally moving down the rest data by one row, and storing the data transmitted in the step 2 into the first row of the storage matrix to update the storage matrix; step 4, performing time accumulation on each distance unit of the time storage matrix, and judging whether each point after accumulation is a clutter boundary point or not; and step 5, obtaining aclutter partitioning result and outputting the result.

The invention discloses a BVI-CFAR target detection algorithm of a variation index based on Bayesian interference control, and the algorithm comprises the steps: firstly dividing a sliding window into a front detection window and a rear detection window, then dividing the front detection window and the rear detection window into two parts, obtaining four detection windows, then employing the variation index to carry out the analysis of the process of the four sliding windows, judging the position of a window where interference possibly exists, then adopting a Bayesian sliding window interference control method, predicting and compensating window interference, and deducing an expression of the detection process of a Bayesian variation index algorithm. According to the invention, the Bayesian sliding window interference control method is applied to the variation index detection process for the first time, a good multi-target detection effect is achieved, compared with a traditional VI-CFAR algorithm, the invention has the advantages that the problem that the detection probability is greatly reduced when targets appear on the two sides of a VI-CFAR detector is solved, and by means of the application of Bayesian interference control, the influence of interference on the detection process is reduced, the detection performance of the radar in a complex environment and a multi-target environment is improved, and a new method is provided for the CFAR algorithm.

The invention discloses a radar anti-interference system. The radar anti-interference system comprises a radar receiver, a frequency mixer, an intermediate frequency amplifier, a radar transmitter, an interference investigation processing unit, a main control chip, a distance tracking unit and a modulator. The interference investigation processing unit comprises an interference investigation module and an interference processing unit. The interference investigation module comprises a CFAR detector and an orthogonal two-channel coherent detector. The interference processing unit comprises a canceller, a Doppler filter and a constant false-alarm rate processor. The output end of the radar receiver is connected with the input end of the interference investigation processing unit and the input end of the distance tracking unit through the frequency mixer and the intermediate frequency amplifier in sequence. The radar anti-interference system is easy to operate and high in reliability, distance signals of sum beams are modified through the distance tracking unit, the interference investigation processing unit is used for investigating and processing interference signals, the target detection distance is prolonged, the anti-interference capacity of radar is enhanced, and the practical application prospect is wide.

The invention belongs to the technical field of passive MIMO radar signal processing, and relates to a passive radar network centralized detection method based on distance weighting. For a passive radar network system with multiple external radiation sources and multiple passive radar receivers under a low signal-to-noise ratio, all preprocessed reference signals and monitoring signals are uploaded to a data processing center, and signals on different propagation paths are endowed with different weights by using implicit distance priori knowledge of a target position; and a signal model is established according to propagation attenuation characteristics of electromagnetic waves in space, and a constant false alarm detector is deduced based on a generalized maximum likelihood ratio test criterion, so that centralized joint detection is carried out. The method has the beneficial effects that the intensity of signals received by each receiver is considered, the intensity difference of each signal path is also considered, and the target detection performance in a passive radar network system is greatly improved under a low signal-to-noise ratio by using a centralized detection method.

The invention discloses a multi-strategy Weibull CFAR detection method based on a change index, relates to the technical field of radar adaptive detection, and provides an RWVI-CFAR detector aiming at the problem that the detection performance of a detector in Weibull distribution clutter is reduced in the prior art, the detector has very small CFAR loss in a uniform Weibull clutter environment, and the detection performance of the RWVI-CFAR detector is improved. The method has the detection performance close to TOS-CFAR in a multi-target environment, has the false alarm control capability superior to TGO-CFAR in a clutter edge environment, and effectively solves the problem that the detection performance of a detector is reduced in Weibull distribution clutter. The RWVI-CFAR detector effectively solves the problem that the detection performance of VI-CFAR is reduced in an environment that interference targets exist on both sides by designing a self-adaptive target rejection algorithm.