35 results about "Cfar detector" patented technology

The invention belongs to the field of radar signal processing, and particularly relates to an improved variable index constant false alarm rate detection algorithm. According to the property of the reference window before and after constant false alarm detection, the corresponding constant false alarm processing method is adaptively selected, and the detection performance of the detector in a multi-target background environment can be effectively improved. Aiming at the situation that a traditional variable index constant false alarm detector (VI-CFAR) has targets in front and back reference windows and the detection performance is poor when a detector selection unit averagely selects a small constant false alarm detector (SO-CFAR), the detector is improved by adopting an adaptive deletion selection small constant false alarm detector (ACSO-CFAR), and when a target exists in a single-side reference window, the detection performance of the detector selection unit is improved by adopting an adaptive deletion selection small constant false alarm detector (ACSO-CFAR). An improved unit average maximum constant false alarm rate detector (CA-CFAR) is selected, and compared with a traditional CA-CFAR, the detection performance is improved under the multi-target background.

The invention discloses a cognitive radar active anti-interference waveform generation method and system based on DQN and MAB, and the method comprises the steps: obtaining a radar interference time-frequency diagram based on a radar transmitting signal and an interference signal; based on the radar interference time-frequency diagram, simulating a radar confrontation process and constructing a transmitted waveform strategy library; solving the waveform strategy library by using a DQN algorithm, and optimizing a solving result by using an MAB algorithm to obtain an optimal transmitted waveform; and a CFAR detector is adopted to detect the optimal emission waveform, and the optimal anti-interference waveform and parameters are generated. According to the method, the global strategy search of the DQN and the local parameter optimization of the MAB are fused, the anti-interference dynamic adaptation problem in a multi-pulse time sequence interaction scene is effectively solved, and the method is suitable for the real-time anti-interference requirements of an airborne radar electronic countermeasure system and a complex electromagnetic environment.

The invention discloses an improved self-adaptive CFAR (Constant False Alarm Rate) detection method for defects of worsted wool fabric, and relates to the technical field of defect detection of worsted wool fabric. The invention relates to an improved self-adaptive CFAR (Constant False Alarm Rate) detection method for defects of worsted wool fabric. The method comprises the following steps: shooting a color image of the worsted wool fabric by adopting a CCD (Charge Coupled Device) camera; converting the acquired color image into a grayscale image to obtain a grayscale value of each pixel; traversing the target grayscale image pixel by pixel by using a sliding window type CFAR detector; wherein the sliding window comprises a front edge sliding window and a rear edge sliding window; the combination of the leading edge sliding window and the trailing edge sliding window is used as a reference unit. According to the scheme, through the double-sliding-window structure design and the mean value self-adaptive selection algorithm, the false detection problem of the light and shade junction area is accurately restrained, and the false detection rate is remarkably reduced; the technical bottleneck that the false detection rate is high due to the fact that the fabric is uneven is broken through, double optimization of the false detection rate and the detection rate is achieved in a complex industrial scene, and a reliable technical scheme is provided for intelligent quality inspection of the worsted fabric.

The invention discloses a radar unsupervised neural network CFAR detection method based on local clutter power estimation, relates to the technical field of radar detection, and aims to solve the problem of low CFAR detection accuracy caused by low local clutter power estimation accuracy of an existing adaptive CFAR detector in a complex non-uniform environment. According to the method, the local clutter power at the to-be-detected unit is accurately estimated by using the neural network, and the technical scheme of the invention prevents background information loss after mask shielding caused by too dense CUT (peak point), thereby improving the prediction accuracy of the local clutter power at the mask, and improving the prediction accuracy of the local clutter power at the to-be-detected unit. And finally, the CFAR detection accuracy under the complex non-uniform background is improved. Moreover, the method avoids the training of real manual annotation data, can achieve the training through the large-scale unannotated data in actual use, and improves the generalization of the method for the measured data compared with an intelligent radar target.

The invention discloses a minimum inverse eigenvalue radar signal detection method adopting stochastic resonance preprocessing, and the method comprises the steps: carrying out the secondary sampling of a received signal, compressing the frequency of the signal to match a stochastic resonance system, and expanding the adaptability of the stochastic resonance system to a higher-frequency signal. Then performing stochastic resonance preprocessing on the signals, enhancing signal energy under a low signal-to-noise ratio to obtain processed signal information, sending the processed signals to a CFAR signal detector, constructing a covariance matrix of the signals of the reference unit and the detection unit, performing eigenvalue decomposition on the covariance matrix to obtain a minimum inverse eigenvalue, and finally obtaining a minimum inverse eigenvalue; the minimum inverse feature value of the detection unit is compared with the average minimum inverse feature value of the reference unit to determine the existence problem of the signal. According to the method, the problems that a matrix constant false alarm detector based on the maximum characteristic value is easily influenced by noise uncertainty and a constant false alarm detector based on information geometry has relatively high calculation complexity are solved.

The invention provides a CA-CFAR detector detection probability analysis method and system under a reverberation background, and belongs to the field of detection probability calculation. The method aims at solving the problems that an existing Monte Carlo simulation method for analyzing the false alarm probability of a detector is large in calculation amount, low in efficiency and lack of theoretical support. According to the method, the calculation model of the detection probability is established based on the Gamma distribution and the PHCSR distribution, the detection probability of the CA-CFAR detector under the reverberation background can be directly calculated, the problems that a traditional Monte Carlo simulation method is large in calculation amount and long in consumed time are solved, the calculation efficiency is remarkably improved, and a theoretical basis is provided for evaluation of the detection probability.

The invention discloses a clustering hybrid accumulation optimization method for multi-unmanned aerial vehicle radar target signal detection, which is applied to the technical field of radar signals, and aims to solve the problems that in the multi-unmanned aerial vehicle radar detection process, part of unmanned aerial vehicles are sparsely distributed in space, full-coherent accumulation cannot be realized, and full-non-coherent accumulation is relatively low in signal-to-noise ratio gain. According to the method, a clustering hybrid accumulation processing architecture (coherent accumulation in clusters and non-coherent accumulation among clusters) is constructed, the calculation amount and hybrid accumulation signal-to-noise ratio gain in the hybrid accumulation process are analyzed and calculated, a clustering hybrid accumulation optimization model is established according to the two indexes, and then a meta-heuristic bidirectional clustering processing method is designed. And thus, an optimized clustering processing result is obtained, effective accumulation of echo signals is realized, and finally, target detection is realized through a CFAR (Constant False Alarm Rate) detector.

The invention provides a non-uniform double-pulse modulation method based on a TCM. The non-uniform double-pulse modulation method comprises the steps that 1, a non-uniform double-pulse period modulation model is constructed; 2, carrying out radar echo modulation analysis; 3, analyzing a pulse compression result; and step 4, regulation and control effect analysis based on a CFAR detector. According to the method, the key parameter of the time delay factor is innovatively introduced, the modulation flexibility and unpredictability are greatly improved through the collaborative design of three degrees of freedom, and the problems of insufficient flexibility and the like existing in a radar target feature regulation and control method are solved. Compared with a regulation and control method based on TCM uniform monopulse period modulation, the generated harmonic peak amplitude distribution is diversified, screening of regular signals by a radar CFAR detector is effectively avoided, and the number of effective harmonic peaks passing detection is remarkably increased. According to the invention, precise suppression or enhancement of specific-order harmonics can be realized, and a technical approach is provided for implementing flexible and dynamic radar target feature regulation and control.

The application discloses a false alarm controllable radar target detection method based on a differential Neumann Pearson criterion and belongs to the technical field of radar target detection. The application aims at the problem that the NP criterion cannot be directly used for neural network training due to non-differentiability in radar target detection based on a neural network. The application comprises the following steps: a traditional CFAR detector is used to process radar AR spectrum original data to extract candidate targets, and sample image blocks centered on the candidate targets are generated; all sample image blocks are used to construct a training data set, and labels are configured for all training data; a small classification network model with controllable false alarm is constructed, network parameters are iteratively optimized and trained by using the training data, and a trained classification network model is obtained; and real-time radar AR spectrum data are used to obtain image blocks to be detected, and the trained classification network model is used for target detection. The application realizes controllable false alarm radar target detection.

The invention provides a three-dimensional CFAR detection method based on semantic segmentation, and belongs to the field of radar target detection. Aiming at performance reduction caused by target amplitude flicker in single-frame RD image detection, a CFAR detector reference window is extended to range, Doppler and frame three dimensions so as to contain more effective samples, a reference unit is selected in combination with semantic information, and the method comprises the following steps: integrating multiple frames of RD images to form a three-dimensional data matrix, then carrying out semantic segmentation on three-dimensional radar data, and obtaining a three-dimensional data matrix; and constructing a three-dimensional semantic sign matrix. In CFAR detection, a reference window clutter unit is eliminated by means of the matrix so as to estimate background power, and finally whether a target exists or not is judged. According to the method, the problem that the detection performance is reduced due to target amplitude flicker in single-frame detection is effectively avoided, the adaptability of a detection algorithm to target fluctuation change is remarkably improved, the calculated amount is small, and engineering implementation is easy.

The invention discloses a vertical measurement ionogram O / X wave separation method based on polarization matched filtering, and the method comprises the steps: obtaining a dual-channel complex signal, and carrying out the incoherent accumulation to obtain echo intensity; using an OS-CFAR detector and morphological processing to generate a mask; calculating a phase difference based on the mask, and fitting through a double-Gaussian mixture model to obtain O-wave and X-wave characteristic phases; estimating a complex polarization ratio; constructing a polarization matched filter weight vector, multiplying the polarization matched filter weight vector with the dual-channel signal, and separating to obtain an O / X wave complex signal matrix; a separated ionogram is generated through amplitude calculation and detection. The method does not depend on circular polarization hypothesis, actual polarization parameters can be directly estimated, high-precision separation is achieved through polarization matched filtering, and robustness and accuracy are improved.

The invention relates to the field of radar signal CFAR processing, in particular to a radar signal CFAR algorithm implementation method based on GPU acceleration. According to the technical scheme, the method comprises the following steps: calculating module values of a main channel signal and an auxiliary channel signal after being subjected to an MTI / MTD algorithm; after signal noise parameter estimation of different sampling points is completed, signal noise parameter estimation data of different sampling points are compared with a signal threshold, and a distance dimension peak value is searched; the maximum value of each sampling point signal in the same distance dimension and different Doppler dimensions is calculated, when two-dimensional CFAR processing is carried out, zero Doppler output is detected through a clutter map, output of each of other Doppler filters is detected through a unit average CFAR detector, and the maximum value of each sampling point signal in the same distance dimension and different Doppler dimensions is calculated. The output of the clutter map detection and the output of the CFAR detector are combined in a large selection mode to serve as a final detection result. The method is suitable for radar signal CFAR processing.

A method for radar signal processing separates an electric signal generated by Doppler processors of a radar into alternating current (AC) magnitudes with Doppler frequencies corresponding to absolute values between 0.1 and 4 Hz and direct current (DC) magnitudes with no Doppler frequency. An AC line and a DC line are generated, based on the AC magnitudes and the DC magnitudes, respectively, by adding respective energies across Doppler cells for each range cell and subsequently applying CFAR detector algorithm respectively. An energy level difference between an energy spike in the AC line and another energy spike in the DC line is calculated and compared with a threshold set for detecting a breathing target that is stationary. If the energy level difference is greater than the threshold, the breathing target that is stationary is detected and reported to the radar operator.

The invention provides a CA-CFAR detector detection probability analysis method and system under a noise background, and belongs to the field of detection probability calculation methods. The objective of the invention is to solve the problem that there is no accurate analytical relationship between the detection probability and the detection threshold of the existing CA-CFAR detector. The method comprises the following steps: fitting probability distribution containing echo signal test statistics by using Gamma distribution, calculating signal amplitude, and calculating detection probability. According to the method, a calculation model of the detection probability is established, and the detection probability of the CA-CFAR detector under the exponential distribution noise background can be directly calculated; the problems that a traditional Monte Carlo simulation method is large in calculation amount and long in consumed time are solved, the calculation efficiency is remarkably improved, and a theoretical basis is provided for evaluation of the detection probability.

The present invention provides a target detection method, device and equipment based on frequency domain Hurst exponent. The method comprises: obtaining a frequency spectrum of a radar echo sequence; extracting frequency domain Hurst exponent features from the frequency spectrum of the radar echo sequence; determining a detection statistic based on the frequency domain Hurst exponent features; generating a decision threshold corresponding to the false alarm probability based on a preset false alarm probability of a constant false alarm detector; determining that a target unit is detected when the detection statistic is greater than or equal to the decision threshold; and determining that the target unit is not detected when the detection statistic is less than the decision threshold. Compared with the related art that uses a likelihood ratio detector to detect targets, the embodiments of the present invention use a constant false alarm detector based on frequency domain Hurst exponent features to detect target units, which can effectively control the false alarm rate and increase the probability of correct detection of target units under a certain false alarm rate.

The present invention discloses a spaceborne pitch-dimensional frequency-scanned beam AMTI method. By introducing time delays into different pitch-dimensional antenna channels of a spaceborne pitch-dimensional frequency-scanned beam AMTI system, the scanning angle of the frequency-scanned beam is greatly increased based on a certain transmit signal bandwidth; range pulse compression is performed on the received echo signals, and then clutter suppression is carried out by utilizing the height-dimensional information difference between ground clutter and airborne moving targets; finally, an airborne moving target can be detected by using a transverse false alarm CFAR detector, and tracking of the airborne moving target is completed in combination with a Kalman filter. The present invention can effectively improve the beam scanning coverage area so as to achieve the effect of wide-area detection of non-cooperative moving targets in the air.

The invention provides a pulse Doppler radar target distance-Doppler feature modulation method based on a phase modulation surface. The pulse Doppler radar target distance-Doppler feature modulation method comprises the steps of 1, presetting a coded modulation waveform; 2, modulating the position characteristics of the false target; 3, speed characteristic modulation of the false target; and step 4, deception of the CFAR detector is carried out. The PSS-based PD radar target distance-Doppler feature modulation method is innovatively provided, and the application of the PSS in the aspect of radar target multi-dimensional feature joint modulation is expanded. According to the invention, the cost is lower, the system complexity is lower, and the electromagnetic concealment is stronger. By applying different coding waveforms to the PSS feature modulation reflector, flexible and diversified feature modulation styles can be generated, the limitation of the traditional PD radar passive feature modulation method on functions of limited effect, insufficient flexibility and the like is overcome, and the advantage of reusability of an active feature modulation technology is also taken into account.

The present invention provides a method and system for detecting the trajectory of a weak moving target based on sonar images. The method includes: using the shape parameter, scale parameter in the gray level and a preset false alarm rate as the input parameters of the CA-CFAR detector; taking a rectangular window as the selection criterion for the reference unit of the target detection sonar image, traversing each pixel value on the target detection sonar image to obtain the accumulated value of the echo intensity; when the accumulated value of the echo intensity is greater than or equal to the preset detection threshold, discarding the trajectory with the minimum energy, and splicing the echo intensities on the remaining trajectories; when the accumulated value of the echo intensity is less than the preset detection threshold, splicing all the trajectories to obtain the corresponding reference unit; estimating the reference unit through maximum likelihood estimation to obtain the shape parameter and scale parameter of the reverberation noise distribution, and obtaining the estimated value of the detection threshold; comparing the detection unit with the estimated value of the detection threshold to obtain the detection result.

The invention provides a sea surface target re-detection method based on deep learning noise level estimation, a storage medium and electronic equipment, and the method comprises the following steps: firstly, calculating a background noise level and a detection threshold value through a VI-CFAR detector, and judging whether a target exists or not according to a binary hypothesis; on the basis of the existence of the target, further judging whether target missing detection exists or not by using a signal-to-noise ratio; secondly, if missing detection is carried out, generating more RD image samples by utilizing a generative adversarial network, and realizing expansion of a data set; inputting the expanded data set into a convolutional neural network, carrying out feature extraction, recognizing and replacing a salient target with environmental noise, and accurately estimating the noise level; and finally, the noise level estimation result is fed back to the VI-CFAR detector, and the detection threshold is recalculated, so that the dynamic adjustment of the detection threshold is realized, and the problem of missing detection in the environment that the weak target is submerged is solved.

The present invention discloses a method for optimizing a statistical distribution model of sea clutter simulation for CFAR detection, belonging to the technical field of sea clutter simulation, and specifically including the following steps: S1, select measured sea clutter data and estimate the parameters of the statistical distribution model; S2, adopt the integral goodness-of-fit test, calculate the integral goodness-of-fit test statistic of each statistical distribution model, and optimize the statistical distribution model that best fits the selected data according to the decision criterion; S3, calculate the spatial and temporal correlation lengths of the intensity of the selected data and generate a two-dimensional autocorrelation matrix; S4, adopt the memory nonlinear transformation method to simulate and generate a spatio-temporally correlated sea clutter sequence driven by the measured statistical parametric model for training the CFAR detector. The present invention adopts the above method for optimizing the statistical distribution model of sea clutter simulation for CFAR detection, optimizes the statistical distribution model with a detection threshold slightly higher than the measured data, and is used to generate a spatio-temporally correlated sea clutter sequence, improving the accuracy and robustness of the CFAR detector.

The invention discloses a radar unsupervised neural network CFAR detection method based on local clutter power estimation, relates to the technical field of radar detection, and aims to solve the problem of low CFAR detection accuracy caused by low local clutter power estimation accuracy of an existing adaptive CFAR detector in a complex non-uniform environment. According to the method, the local clutter power at the to-be-detected unit is accurately estimated by using the neural network, and the technical scheme of the invention prevents background information loss after mask shielding caused by too dense CUT (peak point), thereby improving the prediction accuracy of the local clutter power at the mask, and improving the prediction accuracy of the local clutter power at the to-be-detected unit. And finally, the CFAR detection accuracy under the complex non-uniform background is improved. Moreover, the method avoids the training of real manual annotation data, can achieve the training through the large-scale unannotated data in actual use, and improves the generalization of the method for the measured data compared with an intelligent radar target.

The application discloses a kind of multi-unmanned aerial vehicle radar target signal detection's cluster hybrid accumulation optimization method, applied to radar signal technical field, for multi-unmanned aerial vehicle radar detection process in part, unmanned aerial vehicle spatial distribution sparse, cannot realize full coherent accumulation, and full non-coherent accumulation signal-to-noise ratio gain is lower Problem;The application constructs cluster hybrid accumulation processing architecture (intra-cluster coherent accumulation and inter-cluster non-coherent accumulation), the amount of calculation in the calculation hybrid accumulation process is analyzed with hybrid accumulation signal-to-noise ratio gain, and according to the two indexes, a cluster hybrid accumulation optimization model is established, and then a cluster processing method based on meta-heuristic bidirectional is designed, so as to obtain the optimized cluster processing result and realize the effective accumulation of echo signal, and finally realize target detection through CFAR detector.

The invention relates to a clutter map detection method based on OTSU-CCA, and belongs to the technical field of radar target detection. Comprising the following steps: 1) performing KS normality test on a continuous scanning period sample of a to-be-detected unit, and verifying whether the continuous scanning period sample of the to-be-detected unit obeys exponential distribution or not; 2) sequencing the samples in the continuous scanning period of the unit to be detected according to a sequence from small to large; 3) traversing each sorted sample, and calculating the corresponding between-class variance of the sample; the sorted sample sequence number corresponding to the maximum between-class variance value is the demarcation point of the sample containing the interference target and the sample not containing the interference target; and 4) carrying out detection processing by adopting a CCA-CFAR detector, and further carrying out screening. According to the method, the adaptive screening process is added, so that interference target samples in continuous scanning period samples of the azimuth-distance unit to be detected can be eliminated, robust estimation of the background power level is realized, and the detection performance is improved.

The invention discloses a multi-channel parallel two-dimensional CFAR (Constant False Alarm Rate) detector based on an FPGA (Field Programmable Gate Array), and belongs to the technical field of radar target detection. According to the method, multiple units are detected at the same time in each clock period from data rearrangement and MTD, sliding windows of a data Doppler domain are avoided through multi-channel parallel detection, and the detection speed is effectively increased. Compared with a traditional implementation method for simultaneously carrying out data window sliding in the Doppler direction and the distance dimension direction on the basis of FPGA two-dimensional CFAR, the method has the advantages that multi-channel parallel window sliding and arbitration design are adopted, the reusability of modules is fully considered, two-dimensional CFAR detection of different types and different detection windows can be realized, and the detection efficiency is improved. The implementation time of the two-dimensional CFAR detection can be effectively saved, and the implementation difficulty of the two-dimensional CFAR detection can be reduced.

The present invention discloses a method and system for generating active anti-interference waveforms for cognitive radars based on DQN and MAB, wherein the method comprises: obtaining a radar interference time-frequency diagram based on radar transmission signals and interference signals; simulating a radar countermeasure process and constructing a transmission waveform strategy library based on the radar interference time-frequency diagram; solving the waveform strategy library using a DQN algorithm, and optimizing the solution using a MAB algorithm to obtain an optimal transmission waveform; detecting the optimal transmission waveform using a CFAR detector to generate an optimal anti-interference waveform and parameters. The present invention integrates the global strategy search of DQN with the local parameter optimization of MAB, effectively solving the problem of dynamic anti-interference adaptation in multi-pulse timing interaction scenarios, and is suitable for airborne radar electronic countermeasure systems and real-time anti-interference requirements in complex electromagnetic environments.

This invention discloses a SAR image target detection method based on the fusion of statistical and structural characteristics. It utilizes a CFAR detector and a Faster-RCNN detection network to iteratively identify targets in the SAR image under test until the alternating iteration termination condition is met, thus obtaining the target identification result of the SAR image. This invention fuses the CFAR detector and the Faster-RCNN detection network. Based on the large number of false alarm bounding boxes obtained from the CFAR detector, soft annotations of the bounding boxes are obtained through pixel-level to target-level fusion calculations. The Faster-RCNN detection network is used as the basic training framework, and the bounding boxes output by the Faster-RCNN detection network are used as the protected area size of the CFAR detector, reducing the mixing of target samples with clutter samples and making the parameter estimation of the statistical distribution model more accurate.

The invention discloses a CFAR detection method and system based on multiple screening modules and adaptive weighting, and mainly solves the problems of poor detection performance and difficult configuration in a complex clutter environment in the prior art. The method comprises the following implementation steps of: eliminating other interference signals in a processing window by a phase characteristic screening module; the repeated eliminating module is used for eliminating obvious noise or clutter peaks; the result set establishing module establishes a result set by utilizing environmental characterization VI and MR; the weighted CFAR detector initialization and weighted distribution module is used for carrying out corresponding weighted OS-CFAR detector initialization and weighted distribution on various complex clutter environments; and the detection output module calculates a detection threshold according to the weight and outputs a detection result. According to the method, the weights of the VI and MR parameters are calculated, and the OS-CFAR detectors are subjected to weighted output through the weights, so that the method still has the advantage of keeping relatively high detection performance in various clutter environments while the configuration complexity and the operation complexity are reduced.

The invention discloses a self-adaptive high-frequency ground wave radar target detection method and system, and relates to the technical field of target detection. The target detection method comprises the following steps: performing two-dimensional Fourier transform on high-frequency ground wave radar echoes to obtain a distance Doppler spectrum so as to distinguish a shore-based radar from a shipborne radar; preliminarily detecting a shore-based radar spectrum through a constant false alarm detector to obtain a suspicious target index, traversing the index to divide connected regions, and judging single and group targets; positioning, tracking, category reasoning and tracking optimization are carried out after single-target secondary detection; and after the center of the group target is determined, positioning, tracking, identifying and associating AIS data. And carrying out preliminary detection on a ship-borne radar spectrum, then intercepting an image, compensating and denoising, detecting and determining a target unit index again, positioning, tracking and associating AIS data to obtain a category, predicting a motion direction, adjusting a posture, and continuing tracking. The problem that an existing high-frequency ground wave radar is large in error in the ocean target detection process is solved, and the accuracy of ocean target detection is improved.

The invention relates to an adversarial sample parameter selection method and system based on non-dominated sorting, and belongs to the field of intelligent optimization algorithms and image processing. The method comprises the specific steps of population initialization, objective function value calculation, non-dominated sorting, congestion degree calculation, evolution operation, elite retention, termination judgment and CFAR detector screening. According to the method, the Pareto leading edge can be quickly converged, the parameter space is effectively explored, local optimum is avoided, a group of Pareto optimal solutions are output, and a user can select an optimal parameter combination according to actual requirements. Critical parameters which can be detected to obviously generate adversarial attacks are screened out through a CFAR detector, and the concealment and the attack effect of adversarial samples are further improved.

The invention discloses a two-stage distance and speed estimation method for a frequency agile radar, and the method comprises the steps: carrying out the first-stage linear processing of an echo signal, calculating the square of the amplitude module value of each pulse, carrying out the summation, judging whether a target is an effective target or not at a peak point of a summation result through a constant false alarm detector, and carrying out the second-stage linear processing. Successively detecting new targets and circularly correcting parameters of the detected targets until a detection termination condition is met, and obtaining a final set of distances and amplitudes of all the targets; and performing second-stage linear processing on the echoes, and estimating the accurate distance and speed of each target through matched filtering and a Newton method according to the distance unit of each target obtained in the first stage. The invention also provides a two-stage distance speed estimation device. The method provided by the invention is used for solving the problems of poor precision, high complexity, difficult detection and the like when the frequency agile radar detects the distance and the speed of the target.