Multi-mode wave crest detection method and device applied to SPAD detector and radar equipment

Abstract

The invention provides a multi-mode wave crest detection method and device applied to an SPAD detector and radar equipment, and the method comprises the steps: constructing histogram data, carrying out the signal quality evaluation of the histogram data, obtaining a signal quality parameter, and carrying out the detection of the signal quality parameter; and selecting a target detection mode corresponding to the current ranging scene from preset wave crest detection modes based on the signal quality parameters, and performing data processing on the histogram data according to the target detection mode to obtain a target ranging distance. According to the multi-mode wave crest detection method provided by the invention, signal quality evaluation is carried out on histogram data constructed in an SPAD detector, and a current ranging scene can be accurately diagnosed according to signal quality parameters, so that a corresponding target detection mode is selected; automation of a detection strategy is realized, the labor cost is reduced, and the intelligence is high; and in the actual calculation process, the target detection mode corresponding to the current distance measurement scene can be selected in time, so that the waste of calculation resources is avoided, and the resource use is optimized.

Description

Technical Field

[0001] This invention relates to the fields of photoelectric detection and signal processing technology, and in particular to a multi-mode peak detection method, apparatus and radar equipment for SPAD detectors. Background Technology

[0002] SPAD (Single-Photon Avalanche Diode) detectors possess extremely high sensitivity and are widely used in fields such as lidar and 3D imaging. SPAD constructs a histogram by statistically analyzing the time of flight of photons and calculates the target distance by detecting the echo peaks in the histogram. Detecting the peaks in the histogram using a SPAD detector is crucial.

[0003] Currently, peak detection in histograms often employs a single algorithm, such as: 1) Local Extremum Method: This method compares adjacent points to determine the local maximum value. However, it is highly susceptible to noise and prone to false detections. 2) Centroid Method: This method calculates the signal energy center to determine the peak position. It has good noise resistance but low resolution in cases with multiple peaks. 3) Fixed Threshold Method: This method sets a fixed threshold to determine the peak, but it cannot dynamically change according to the signal intensity.

[0004] Therefore, it is necessary to provide a peak detection algorithm that can adapt to multiple signal quality detection environments to ensure its robustness when signal strength and noise levels change drastically. Summary of the Invention

[0005] This invention provides a multi-mode peak detection method, apparatus, and radar equipment for SPAD detectors, to solve the technical problems of the above-mentioned peak detection methods being greatly affected by noise, having low resolution, and poor adaptability.

[0006] In a first aspect, the present invention provides a multi-mode peak detection method for SPAD detectors, comprising: Construct histogram data; The histogram data is subjected to signal quality assessment to obtain signal quality parameters, which include signal-to-noise ratio, total effective photon count, peak significance, and waveform entropy. Based on multiple signal quality parameters, a target detection mode corresponding to the current ranging scenario is selected from preset peak detection modes. Based on the target detection mode, the histogram data is processed to obtain the target ranging distance.

[0007] In one embodiment of the present invention, signal quality assessment is performed on the histogram data to obtain signal quality parameters, including: smoothing and filtering the histogram data to obtain a preprocessed histogram; identifying each candidate sampled signal in the preprocessed histogram and its corresponding photon time-of-flight index, and obtaining the noise floor of the preprocessed histogram; determining the signal-to-noise ratio based on the maximum photon count in the candidate sampled signals and the noise floor; determining the total effective photon count based on the photon counts of multiple candidate sampled signals and the average noise value in the noise floor; identifying local peaks from multiple candidate sampled signals, calculating the ratio of the average intensity of each local peak to the average intensity of its two adjacent valleys, and taking the maximum value of all ratios as the peak significance; normalizing the photon counts of the preprocessed histogram, and determining the waveform entropy based on the normalized probability distribution function.

[0008] In one embodiment of the present invention, the preset peak detection mode includes a first mode, a second mode, and a third mode. Selecting a target peak detection mode corresponding to the current ranging scenario from the preset peak detection modes based on multiple signal quality parameters includes: when the signal-to-noise ratio (SNR) is greater than a first SNR threshold, and the peak saliency is greater than a preset peak threshold, while the waveform entropy is less than a waveform entropy threshold, the first mode is selected as the target peak detection mode; when the SNR is less than or equal to the first SNR threshold but greater than a second SNR threshold, and the total effective photon count is greater than the total effective light... When the waveform entropy is less than the waveform entropy threshold, the second mode is used as the target peak detection mode; when the signal-to-noise ratio is less than or equal to the second signal-to-noise ratio threshold, or the peak saliency is less than or equal to the preset peak threshold, or the total effective photon count is less than or equal to the total effective photon count threshold, or the waveform entropy is greater than or equal to the waveform entropy threshold, the third mode is used as the target peak detection mode; wherein, the first mode is suitable for strong signal scenarios, the second mode is suitable for weak signal scenarios, and the third mode is suitable for high noise scenarios.

[0009] In one embodiment of the present invention, when the target detection mode is a first mode, data processing is performed on the histogram data based on the target detection mode to obtain the target ranging distance, including: performing global and local maximum indexing on the preprocessed histogram, and taking the maximum value of the photon count greater than the noise floor as the first mode peak; sorting multiple first mode peaks in descending order based on the photon count, and determining the corresponding number of target ranging distances based on the photon time-of-flight index of the first preset number of first mode peaks.

[0010] In one embodiment of the present invention, when the target detection mode is the second mode, data processing is performed on the histogram data based on the target detection mode to obtain the target ranging distance, including: determining the second mode peak based on the preprocessed histogram, wherein the second mode peak is the highest peak point; dividing the peak region based on a preset width rule with the position of the second mode peak as the center; determining the centroid position of the second mode peak within the peak region based on the photon time-of-flight index and its corresponding photon count, so as to output the target ranging distance according to the centroid position.

[0011] In one embodiment of the present invention, when the target detection mode is the third mode, data processing is performed on the histogram data based on the target detection mode to obtain the target ranging distance, including: performing impulse response filtering on the preprocessed histogram based on a preset cutoff frequency to obtain an enhanced smooth waveform; determining a noise detection threshold based on the noise floor; identifying regions in the enhanced smooth waveform where the photon count exceeds the noise detection threshold as valid signals, and determining the peak value of the valid signal, and outputting the target ranging distance based on the photon time-of-flight index corresponding to the peak value.

[0012] In one embodiment of the present invention, the peak detection method further includes: performing a comprehensive score on the current ranging scene based on the signal quality parameters to obtain a scene score; and switching the target detection mode based on a preset scene threshold when the scene score changes.

[0013] In one embodiment of the present invention, the preset scene threshold includes a first scene threshold and a second scene threshold. When the scene score changes, the target detection mode is switched based on the preset scene threshold, including: when the direction of change of the target detection mode is from the first mode to the third mode, the target detection mode is switched based on the first scene threshold; when the direction of change of the target detection mode is from the third mode to the first mode, the target detection mode is switched based on the second scene threshold; wherein, the first scene threshold is greater than or equal to the second scene threshold.

[0014] Secondly, the present invention also provides a multi-mode peak detection device for SPAD detectors, the device comprising: The data construction module is used to build histogram data; The signal quality assessment module is used to assess the signal quality of the histogram data and obtain signal quality parameters, including signal-to-noise ratio, total effective photon count, peak significance, and waveform entropy. The mode selection module is used to select a target detection mode corresponding to the current ranging scenario from a preset peak detection mode based on multiple signal quality parameters. The distance determination module is used to process the histogram data based on the target detection mode to obtain the target distance.

[0015] Thirdly, the present invention also provides a radar device, including a multi-mode peak detection device for SPAD detectors as described above.

[0016] The beneficial effects of this invention are as follows: This invention proposes a multi-mode peak detection method, device, and radar equipment for SPAD detectors. The method includes: constructing histogram data; evaluating the signal quality of the histogram data to obtain signal quality parameters; selecting a target detection mode corresponding to the current ranging scenario from preset peak detection modes based on the signal quality parameters; and processing the histogram data according to the target detection mode to obtain the target ranging distance. This invention provides a multi-mode peak detection method that evaluates the signal quality of the histogram data constructed within the SPAD detector. It can accurately diagnose the current ranging scenario based on the signal quality parameters, thereby selecting the corresponding target detection mode. It automates the detection strategy, reduces labor costs, and is highly intelligent. Furthermore, in the actual calculation process, it can promptly select the target detection mode corresponding to the current ranging scenario, eliminating waste of computational resources and optimizing resource utilization. Attached Figure Description

[0017] The accompanying drawings, which are incorporated in and form part of this specification, illustrate embodiments consistent with this application and, together with the description, serve to explain the principles of this application. It is obvious that the drawings described below are merely some embodiments of this application, and those skilled in the art can obtain other drawings based on these drawings without any inventive effort.

[0018] In the attached diagram: Figure 1 This is a flowchart illustrating the multi-mode peak detection method applied to a SPAD detector provided in an embodiment of the present invention. Figure 2 This is a flowchart illustrating the multi-mode peak detection method for SPAD detectors provided in this embodiment of the invention. Figure 3 These are histogram data for different ranging scenarios provided in the embodiments of the present invention; Figure 4 This is a block diagram of a multi-mode peak detection device for SPAD detectors provided in an embodiment of the present invention. Detailed Implementation

[0019] The following specific examples illustrate the implementation of the present invention. Those skilled in the art can easily understand other advantages and effects of the present invention from the content disclosed in this specification. The present invention can also be implemented or applied through other different specific embodiments. Various details in this specification can also be modified or changed based on different viewpoints and applications without departing from the spirit of the present invention. In the absence of conflict, the following embodiments and features in the embodiments can be combined with each other.

[0020] It should be noted that the illustrations provided in the following embodiments are only schematic representations of the basic concept of the present invention. The drawings only show the components related to the present invention and are not drawn according to the actual number, shape and size of the components in the actual implementation. In the actual implementation, the form, quantity and proportion of each component can be arbitrarily changed, and the layout of the components may also be more complex.

[0021] In the following description, numerous details are explored to provide a more thorough explanation of embodiments of the invention. However, it will be apparent to those skilled in the art that embodiments of the invention may be practiced without these specific details. In other embodiments, well-known structures and devices are shown in block diagram form rather than in detail to avoid obscuring embodiments of the invention.

[0022] To solve the technical problems mentioned in the background section, such as Figure 1 As shown, this application provides a multi-mode peak detection method for SPAD detectors, which includes at least steps S110 to S140: S110. Construct histogram data.

[0023] Specifically, such as Figure 2 As shown, in step S201, the optical signal is input into the SPAD detector to form histogram data. The SPAD detector is a complete detection system. Its core is to use a single photon avalanche diode as a photosensitive pixel unit to detect extremely weak light signals (even a single photon), and to process the histogram data constructed from the detected light signal to output the corresponding ranging result.

[0024] S120. Perform signal quality assessment on the histogram data to obtain signal quality parameters, including signal-to-noise ratio, total effective photon count, peak significance, and waveform entropy.

[0025] In detail, signal quality assessment is performed on the histogram data to obtain signal quality parameters, including: smoothing and filtering the histogram data to obtain a preprocessed histogram; identifying each candidate sampled signal in the preprocessed histogram and its corresponding photon time-of-flight index, and obtaining the noise floor of the preprocessed histogram; determining the signal-to-noise ratio based on the maximum photon count in the candidate sampled signals and the noise floor; determining the total effective photon count based on the photon counts of multiple candidate sampled signals and the average noise in the noise floor; identifying local peaks from multiple candidate sampled signals, calculating the ratio of the average intensity of each local peak to the average intensity of its two adjacent valleys, and taking the maximum value of all ratios as the peak significance; normalizing the photon counts in the preprocessed histogram, and determining the waveform entropy based on the normalized probability distribution function.

[0026] Specifically, such as Figure 2 As shown, signal quality is evaluated on the histogram data through steps S202 to S206 to obtain signal quality parameters, including signal-to-noise ratio (SNR), total effective photon count (N_total), peak significance (Max_Salience), and waveform entropy (Entropy). In step S202, the histogram data is preprocessed, which may include an FIR smoothing filter to suppress random noise fluctuations, resulting in a preprocessed histogram. Each candidate sampled signal and its corresponding photon time-of-flight index are identified in the preprocessed histogram, and noise is extracted from the signal blank areas in the preprocessed histogram to obtain a noise floor, which includes the noise mean μ_noise and the noise standard deviation σ_noise. In step S203, the signal-to-noise ratio (SNR) is calculated based on the maximum photon count P_max, the noise mean μ_noise, and the noise standard deviation σ_noise in the candidate sampled signals.

[0027] An FIR smoothing filter is a special type of digital filter whose core purpose is to "smooth" or "average" a signal in order to suppress random noise, remove high-frequency glitches, and thus extract the overall trend or slowly changing components of the signal. It achieves this goal by calculating a weighted average of the signal over a small moving window.

[0028] The expression for determining the signal-to-noise ratio (SNR) is shown in (1): SNR=(P_max-μ_noise) / σ_noise(1) Where SNR is the signal-to-noise ratio; P_max is the maximum photon count in the candidate sampled signal; μ_noise is the average noise value; and σ_noise is the standard deviation of the noise.

[0029] In step S204, the total effective photon count N_total is calculated based on the photon count and noise average μ_noise of multiple candidate sampled signals. The expression for determining the total effective photon count N_total is shown in (2): N_total= (2) Where N_total is the total effective photon count. The photon count is the number of the candidate sampled signal corresponding to the i-th photon time-of-flight index. The noise average value is denoted by , where i is a positive integer and 0 < i ≤ the total photon flight time index.

[0030] In step S205, local peaks of multiple candidate sampling signals are identified, multiple local peaks are obtained, the average intensity of the two valleys adjacent to each local peak is calculated, each local peak is compared with its corresponding average intensity, multiple ratio results are obtained, and the largest ratio result is taken as the peak significance Max_Salience.

[0031] In step S206, the photon counts of all candidate sampled signals in the preprocessed histogram are normalized to a probability distribution function, and the waveform entropy is calculated based on the probability distribution function. The expression for determining the waveform entropy is shown in (3): Entropy=- (3) Where Entropy is the waveform entropy, P(i) is the probability distribution function, and i is the photon flight time index.

[0032] S130. Select the target detection mode corresponding to the current ranging scenario from the preset peak detection modes based on the signal quality parameters.

[0033] In detail, the preset peak detection modes include a first mode, a second mode, and a third mode. Based on signal quality parameters, a target peak detection mode corresponding to the current ranging scenario is selected from the preset peak detection modes, including: when the signal-to-noise ratio (SNR) is greater than a first SNR threshold, and the peak significance is greater than a preset peak threshold, while the waveform entropy is less than a waveform entropy threshold, the first mode is used as the target peak detection mode; when the SNR is less than or equal to the first SNR threshold but greater than the second SNR threshold, and the total effective photon count is greater than the total effective photon count threshold, while the waveform entropy is less than a waveform entropy threshold, the second mode is used as the target peak detection mode; when the SNR is less than or equal to the second SNR threshold, or the peak significance is less than or equal to the preset peak threshold, or the total effective photon count is less than or equal to the total effective photon count threshold, or the waveform entropy is greater than or equal to the waveform entropy threshold, the third mode is used as the target peak detection mode; wherein, the first mode is suitable for strong signal scenarios, the second mode is suitable for weak signal scenarios, and the third mode is suitable for high noise scenarios.

[0034] Specifically, such as Figure 2 As shown, in step S207, feature fusion and decision-making are performed based on signal quality parameters to select a target peak detection mode corresponding to the current ranging scenario. The preset peak detection modes include a first mode, a second mode, and a third mode. The first mode is for scenarios where the current ranging scenario has strong signal strength and object reflection signals are easy to collect, suitable for strong signal scenarios, such as indoor signal detection. The second mode is for scenarios where the current ranging scenario has moderate signal strength and object reflection signal collection conditions are average, suitable for weak signal scenarios, such as outdoor low-reflection distant target scenarios. The third mode is for scenarios where the current ranging scenario has weak signal strength, low visibility, and difficult object reflection signal collection conditions, suitable for high-noise scenarios, such as ranging in foggy or hazy weather.

[0035] like Figure 2As shown, in step 208, based on the signal quality parameter data, the target peak detection mode for the current ranging scenario is selected. When the signal-to-noise ratio (SNR) is greater than the first SNR threshold (20dB, in decibels), the peak salience (Max_Salience) is greater than the preset peak threshold (2.0), and the waveform entropy (Entropy) is less than the waveform entropy threshold (3.0), the first mode is selected as the target peak detection mode corresponding to the current ranging scenario. If this condition is not met, the judgment continues. When the SNR is less than or equal to the first SNR threshold (20dB) but greater than the second SNR threshold (9dB), and the total effective photon count (N_tota) is greater than the first SNR threshold (20dB), the target peak detection mode for the current ranging scenario is selected as the target peak detection mode for the current ranging scenario. If l is greater than the total effective photon count threshold (300) and the waveform entropy is less than the waveform entropy threshold (3.0), the second mode is selected as the target peak detection mode corresponding to the current ranging scenario. If the first two conditions are not met, the judgment continues. If the signal-to-noise ratio (SNR) is less than or equal to the second SNR threshold (9dB), or the peak salience (Max_Salience) is less than or equal to the preset peak threshold (2.0), or the total effective photon count is less than or equal to the total effective photon count threshold (300), or the waveform entropy is greater than or equal to the waveform entropy threshold (3.0), the third mode is selected as the target peak detection mode corresponding to the current ranging scenario.

[0036] S140. Based on the target detection mode, perform data processing on the histogram data to obtain the target ranging distance.

[0037] In detail, when the target detection mode is the first mode, the histogram data is processed based on the target detection mode to obtain the target ranging distance, including: performing global and local maximum indexing on the preprocessed histogram, and taking the maximum value of photon count greater than the noise floor as the first mode peak; sorting multiple first mode peaks in descending order based on photon count, and determining the corresponding number of target ranging distances based on the photon time-of-flight index of the first preset number of first mode peaks.

[0038] Specifically, such as Figure 2 As shown, the target detection mode has been determined to be the first mode. In step S209, a global and local maximum search is performed on the preprocessed histogram. The peak with a photon count greater than the maximum value in the noise floor is taken as the first mode peak. Based on the photon count of each first mode peak, multiple first mode peaks are sorted in descending order to obtain a sequence of first mode peaks sorted from largest to smallest. In step S210, the first 4 first mode peaks are identified. In step S215, the distance to multiple targets is calculated using the photon time-of-flight index of each first mode peak.

[0039] In detail, when the target detection mode is the second mode, the histogram data is processed based on the target detection mode to obtain the target ranging distance, including: determining the second mode peak based on the preprocessed histogram, where the second mode peak is the highest point; dividing the second mode peak into regions based on a preset width rule with the position of the second mode peak as the center; and determining the centroid position of the second mode peak within the peak region based on the photon time-of-flight index and its corresponding photon count, so as to output the target ranging distance based on the centroid.

[0040] Specifically, such as Figure 2 As shown, the target detection mode has been determined to be the second mode. In step S211, the highest peak point is found in the preprocessing histogram to determine the peak of the second mode. Half the height of the peak of the second mode is calculated. A horizontal line is drawn from this height on the peak of the second mode, and the two points where the line intersects the waveform of the peak of the second mode are determined to determine the peak region of the peak of the second mode. In step S212, within the peak region, the centroid position of the peak of the second mode is determined based on the photon time-of-flight index j and the photon count G[j] corresponding to the photon time-of-flight index j. The centroid position is used as the precise peak position of the peak of the second mode, and the target ranging distance is output according to the peak position of the peak of the second mode.

[0041] The expression for determining the centroid position is shown in (4): Centroid=Σ(j*G[j]) / ΣG[j](4) Where Centroid is the centroid position; j is the photon time-of-flight index corresponding to the second mode peak, j>0; G[i] is the photon count at the j-th photon time-of-flight index.

[0042] In detail, when the target detection mode is the third mode, the histogram data is processed based on the target detection mode to obtain the target ranging distance, including: performing impulse response filtering on the preprocessed histogram based on a preset cutoff frequency to obtain an enhanced smooth waveform; determining the noise detection threshold based on the noise floor; identifying the region in the enhanced smooth waveform where the photon count exceeds the noise detection threshold as an effective signal, and determining the peak value of the effective signal, and outputting the target ranging distance based on the photon time-of-flight index corresponding to the peak value.

[0043] Specifically, such as Figure 2 As shown, the target detection mode has been determined to be the third mode. In step S213, an FIR filter with a steeper cutoff frequency is used to perform impulse response filtering on the preprocessed histogram to achieve strong smoothing and obtain an enhanced smooth waveform. The noise detection threshold is calculated by the noise average value and the noise standard deviation. The expression for determining the noise detection threshold is shown in (5): Threshold=μ_noise+3*σ_noise (5) Where Threshold is the noise detection threshold; μ_noise is the average noise value; and σ_noise is the noise standard deviation.

[0044] The region in the enhanced smooth waveform where the photon count exceeds the noise detection threshold is identified as a valid signal, and the peak value of the valid signal is identified. The target ranging distance is output based on the photon time-of-flight index corresponding to the peak value.

[0045] In detail, the peak detection method further includes: comprehensively scoring the current ranging scene based on signal quality parameters to obtain a scene score; and switching the target detection mode based on a preset scene threshold when the scene score changes. Specifically, the peak detection method also includes comprehensively scoring the current ranging scene using signal quality parameters to obtain a scene score for the current ranging scene; however, in actual measurement processes, for example, in foggy weather, the scene score may change as the weather improves. To avoid frequent mode oscillations during mode selection, a preset scene threshold is introduced during the switching between two modes.

[0046] More specifically, the preset scene thresholds include a first scene threshold and a second scene threshold. When the scene score changes, the target detection mode is switched based on the preset scene thresholds, including: when the target detection mode changes from the first mode to the third mode, the target detection mode is switched based on the first scene threshold; when the target detection mode changes from the third mode to the first mode, the target detection mode is switched based on the second scene threshold; wherein, the first scene threshold is greater than or equal to the second scene threshold.

[0047] Specifically, the preset scene thresholds include a first scene threshold and a second scene threshold, with the first scene threshold being greater than the second scene threshold. A scene score of 70 to 100 is suitable for the first mode, a scene score of 40 to 69 is suitable for the second mode, and a scene score of 0 to 39 is suitable for the third mode. When the scene score changes from high to low, i.e., the target detection mode switches from the first mode to the second mode, the second mode to the third mode, or the first mode to the third mode, the speed of scene switching is controlled based on the first scene threshold. For example, a mode switch is only performed when the target detection mode is within the corresponding scene score for 5 consecutive frames. When the scene score changes from low to high, i.e., the target detection mode switches from the third mode to the second mode, the second mode to the first mode, or the third mode to the first mode, the speed of scene switching is controlled based on the second scene threshold. For example, a mode switch is only performed when the target detection mode is within the corresponding scene score for 2 consecutive frames.

[0048] The preset scene thresholds include a first scene threshold and a second scene threshold, with the first scene threshold equal to the second scene threshold. A scene score of 70 to 100 is suitable for the first mode, a scene score of 40 to 69 is suitable for the second mode, and a scene score of 0 to 39 is suitable for the third mode. When the scene score changes from high to low, that is, when the target detection mode switches between the first mode, the second mode, and the third mode, the speed of scene switching is controlled based on the first scene threshold. For example, if it is necessary to be in the corresponding scene score for 3 consecutive frames, then mode switching is performed.

[0049] like Figure 3 As shown, Figure 3 Histogram data provided in this application for different scenarios; such as Figure 3 As shown, the horizontal axis represents 1024 time points, i.e., photon time-of-flight indices, indicating the distribution of flight time on the time axis, and the vertical axis represents photon count. Taking a SPAD detector with histogram data of 1024 photon time-of-flight indices as an example, the temporal resolution of the ranging histogram is set to 100 ps / bin (approximately 1.5 cm distance resolution), the ranging range is 0~15.36 meters (1024 × 0.015 m), the first signal-to-noise ratio threshold is set to 20 dB, the second signal-to-noise ratio threshold is set to 9 dB, the total effective photon count threshold is set to 300, the waveform entropy threshold is set to 3.0, and the preset peak threshold is set to 2.0.

[0050] 1) In a scenario involving strong signals from multiple targets within an indoor space, with a distance of 3 meters from a wall and 5 meters from furniture, the histogram data shows that the photon count at the 200th photon time-of-flight index is 850 (see attached). Figure 3 As shown in 301), the photon count of the secondary echo peak occurring at the 333rd photon time-of-flight index is 420 (as shown in the attached image). Figure 3As shown in Figure 302, the histogram data is preprocessed to obtain a preprocessed histogram. The noise floor is extracted from the preprocessed histogram, with the average noise μ_noise=12 and the standard deviation of noise σ_noise=4. The signal quality is evaluated based on the preprocessed histogram, with a signal-to-noise ratio SNR=(850-12) / 4=209.5 (equal to 23dB), a peak significance Max_Salience=850 / [(15+18) / 2]=51.5, where 15 and 18 are the photon counts at the troughs on both sides of the photon count of 850. The total effective photon count N_total is 1800, and the waveform entropy is 1.2. In this case, SNR > 20dB, Max_Salience > 2.0, and Entropy < 3.0; the first mode is selected to process the peaks in the preprocessed histogram, that is: by performing global and local extreme value search on the preprocessed histogram, two first mode peaks located at the 200th and 333rd photon time-of-flight indices are found, and the time points of the two first mode peaks, 200 and 300, are output, with corresponding target ranging distances of 3 meters and 5 meters, respectively.

[0051] 2) When detecting a black vehicle at 15 meters outside the window, a distant target with low reflectivity, the histogram data shows: a broad, low-amplitude photon count of 65 appears at the 1000th photon time-of-flight index (e.g., ...). Figure 3As shown in Figure 303), the peak shape is diffused, and the full width at half maximum (FWHM) of the peak is approximately 15 photon time-of-flight indices. The histogram data is preprocessed to obtain a preprocessed histogram. The noise floor is extracted from the preprocessed histogram, with a noise mean μ_noise=18 and a noise standard deviation σ_noise=5. The signal quality is evaluated based on the preprocessed histogram, with a signal-to-noise ratio (SNR) of (65-18) / 5=9.4 (equivalent to 9.7dB), a peak significance of Max_Salience=65 / [(22+25) / 2]=2.76, where 22 and 25 are the photon counts at the troughs on both sides of the photon count of 65. The total effective photon count N_total is 320, and the waveform entropy is 2.8. In this case, 9dB < SNR ≤ 20dB, N_total > 300, and Entropy < 3; the second mode is selected to process the peaks in the preprocessed histogram, that is: first, half-wave detection is performed on the preprocessed histogram, and at the 1000th photon time-of-flight index, half the height of the second mode peak, 32.5, is found. Based on this height, half of the peak is truncated, so the peak region of the second mode peak is between the 985th and 1015th photon time-of-flight indices. The centroid is calculated within the peak region, and the centroid position is obtained as the 1002.3rd photon time-of-flight index. The peak position of the second mode peak is output as the 1002.3rd photon time-of-flight index, and the corresponding target ranging distance is 15.034 meters.

[0052] 3) In a high-noise measurement scenario involving a pedestrian at a distance of 8 meters during foggy weather, the histogram data shows the following: the entire histogram data exhibits high floor noise (photon count fluctuates between 20 and 25), with a slight bulge at the 533rd photon time-of-flight index showing a count of 48 (e.g., ...). Figure 3As shown in Figure 304), there is no obvious peak. The histogram data is preprocessed to obtain a preprocessed histogram. The noise floor is extracted from the preprocessed histogram. The noise mean μ_noise=28 and the noise standard deviation σ_noise=8. The signal quality is evaluated on the preprocessed histogram. The signal-to-noise ratio SNR=(48-28) / 5=2.5 (equal to 4dB). The peak significance Max_Salience=48 / [(32+35) / 2]=1.43, where 32 and 35 are the photon counts of the troughs on both sides of the photon count of 48. The total effective photon count N_total is 150 and the waveform entropy is 4.2. In this case, SNR < 9dB, or N_total ≤ 300, or Max_Salience ≤ 2.0, or Entropy ≥ 3; select the third mode to process the peaks in the preprocessed histogram, that is: first, perform enhanced FIR filtering on the preprocessed histogram to achieve smoothing. After filtering, reduce data fluctuations. Calculate the noise detection threshold = μ_noise + 3 × σ_noise = 28 + 3 × 8 = 52. The region greater than the noise detection threshold is taken as the valid signal. The region exceeding the noise detection threshold is between the 533rd photon time-of-flight index and the 533rd photon time-of-flight index. Thus, search for the photon time-of-flight index corresponding to the maximum peak value as the 533rd, and output the peak position at the 533rd photon time-of-flight index, whose corresponding target ranging distance is 8 meters.

[0053] like Figure 4 As shown, the present invention also provides a multi-mode peak detection device for SPAD detectors, the device comprising: Data construction module 410 is used to construct histogram data based on the light signal detected by the SPAD photodiode pixel unit in the SPAD detector; The signal quality assessment module 420 is used to assess the signal quality of histogram data and obtain signal quality parameters, including signal-to-noise ratio, total effective photon count, peak significance, and waveform entropy. The mode selection module 430 is used to select the target detection mode corresponding to the current ranging scenario from the preset peak detection modes based on the signal quality parameters. The distance determination module 440 is used to process the histogram data based on the target detection mode to obtain the target distance.

[0054] It should be noted that the multi-mode peak detection device for SPAD detectors provided in the above embodiments and the multi-mode peak detection method for SPAD detectors provided in the above embodiments belong to the same concept. The specific way of performing each step has been described in detail in the method embodiments, and will not be repeated here.

[0055] This application also provides a radar device including a multi-mode peak detection device for SPAD detectors as described above, to improve the detection accuracy and applicable scenarios of the radar.

[0056] This invention proposes a multi-mode peak detection method, apparatus, and radar equipment for SPAD detectors. The method includes: constructing histogram data; evaluating the signal quality of the histogram data to obtain signal quality parameters; selecting a target detection mode corresponding to the current ranging scenario from preset peak detection modes based on the signal quality parameters; processing the histogram data according to the target detection mode to obtain the target ranging distance; and switching the detection mode based on a preset switching threshold when the current ranging scenario changes. This invention provides a multi-mode peak detection method that evaluates the signal quality of constructed histogram data within the SPAD detector, accurately diagnoses the current ranging scenario based on signal quality parameters, and selects the corresponding target detection mode. It automates the detection strategy, reduces labor costs, and is highly intelligent. Furthermore, during actual calculations, it can promptly select the target detection mode corresponding to the current ranging scenario, eliminating wasted computational resources and optimizing resource utilization. When the testing environment changes, it not only sets a switching threshold but also quickly switches to the corresponding testing mode, improving the robustness of ranging detection.

[0057] The above embodiments are merely illustrative of the principles and effects of the present invention and are not intended to limit the invention. Any person skilled in the art can modify or alter the above embodiments without departing from the spirit and scope of the present invention. Therefore, all equivalent modifications or alterations made by those skilled in the art without departing from the spirit and technical concept disclosed in the present invention should still be covered by the claims of the present invention.

Claims

1. A multi-mode peak detection method applied to SPAD detectors, characterized in that, include: Construct histogram data; The histogram data is subjected to signal quality assessment to obtain signal quality parameters, which include signal-to-noise ratio, total effective photon count, peak significance, and waveform entropy. Based on multiple signal quality parameters, a target detection mode corresponding to the current ranging scenario is selected from preset peak detection modes. Based on the target detection mode, the histogram data is processed to obtain the target ranging distance.

2. The multi-mode peak detection method for SPAD detectors according to claim 1, characterized in that, Signal quality is evaluated using the histogram data to obtain signal quality parameters, including: The histogram data is smoothed and filtered to obtain a preprocessed histogram; Identify each candidate sampled signal in the preprocessed histogram and its corresponding photon time-of-flight index, and obtain the noise floor of the preprocessed histogram; The signal-to-noise ratio is determined based on the maximum photon count in the candidate sampled signal and the noise floor. The total effective photon count is determined based on the photon counts of the multiple candidate sampled signals and the average noise value in the noise substrate; Local peaks are identified from multiple candidate sampling signals, the ratio of the average intensity of each local peak to the average intensity of its two adjacent valleys is calculated, and the maximum value of all ratios is taken as the peak significance. The photon counts of the preprocessed histogram are normalized, and the waveform entropy is determined based on the normalized probability distribution function.

3. The multi-mode peak detection method for SPAD detectors according to claim 2, characterized in that, The preset peak detection modes include a first mode, a second mode, and a third mode. Based on multiple signal quality parameters, a target peak detection mode corresponding to the current ranging scenario is selected from the preset peak detection modes, including: When the signal-to-noise ratio is greater than the first signal-to-noise ratio threshold, and the peak significance is greater than the preset peak threshold, while the waveform entropy is less than the waveform entropy threshold, the first mode is used as the target peak detection mode. When the signal-to-noise ratio is less than or equal to the first signal-to-noise ratio threshold but greater than the second signal-to-noise ratio threshold, and the total effective photon count is greater than the total effective photon count threshold, and the waveform entropy is less than the waveform entropy threshold, the second mode is used as the target peak detection mode. When the signal-to-noise ratio is less than or equal to the second signal-to-noise ratio threshold, or the peak significance is less than or equal to the preset peak threshold, or the total effective photon count is less than or equal to the total effective photon count threshold, or the waveform entropy is greater than or equal to the waveform entropy threshold, the third mode is used as the target peak detection mode. The first mode is suitable for strong signal scenarios, the second mode is suitable for weak signal scenarios, and the third mode is suitable for high noise scenarios.

4. The multi-mode peak detection method for SPAD detectors according to claim 3, characterized in that, When the target detection mode is the first mode, the histogram data is processed based on the target detection mode to obtain the target ranging distance, including: Global and local maxima indexing is performed on the preprocessed histogram, and the maximum value where the photon count is greater than the noise floor is taken as the first mode peak; Based on the photon count, the multiple first mode peaks are sorted in descending order, and the corresponding number of target ranging distances are determined based on the photon time-of-flight index of the first preset number of first mode peaks.

5. The multi-mode peak detection method for SPAD detectors according to claim 3, characterized in that, When the target detection mode is the second mode, the histogram data is processed based on the target detection mode to obtain the target ranging distance, including: The second mode peak is determined based on the preprocessed histogram, and the second mode peak is the highest point of the peak value. Centered on the position of the second mode peak, the peak region is divided according to a preset width rule; Within the peak region, the centroid position of the second mode peak is determined based on the photon time-of-flight index and its corresponding photon count, so as to output the target ranging distance according to the centroid position.

6. The multi-mode peak detection method for SPAD detectors according to claim 3, characterized in that, When the target detection mode is the third mode, the histogram data is processed based on the target detection mode to obtain the target ranging distance, including: The preprocessed histogram is subjected to impulse response filtering based on a preset cutoff frequency to obtain an enhanced smooth waveform. The noise detection threshold is determined based on the noise floor. The region in the enhanced smooth waveform where the photon count exceeds the noise detection threshold is identified as a valid signal, and the peak value of the valid signal is determined, so as to output the target ranging distance based on the photon time-of-flight index corresponding to the peak value.

7. The multi-mode peak detection method for SPAD detectors according to claim 1, characterized in that, The peak detection method further includes: A scene score is obtained by comprehensively evaluating the current ranging scene based on the signal quality parameters. When the scene score changes, the target detection mode is switched based on a preset scene threshold.

8. The multi-mode peak detection method for SPAD detectors according to claim 1, characterized in that, The preset scene thresholds include a first scene threshold and a second scene threshold. When the scene score changes, the target detection mode is switched based on the preset scene thresholds, including: When the target detection mode changes from the first mode to the third mode, the target detection mode is switched based on the first scene threshold. When the target detection mode changes from the third mode to the first mode, the target detection mode is switched based on the second scene threshold. Wherein, the first scenario threshold is greater than or equal to the second scenario threshold.

9. A multi-mode peak detection device for SPAD detectors, characterized in that, The device includes: The data construction module is used to build histogram data; The signal quality assessment module is used to assess the signal quality of the histogram data and obtain signal quality parameters, including signal-to-noise ratio, total effective photon count, peak significance, and waveform entropy. The mode selection module is used to select a target detection mode corresponding to the current ranging scenario from a preset peak detection mode based on multiple signal quality parameters. The distance determination module is used to process the histogram data based on the target detection mode to obtain the target distance.

10. A radar device, characterized in that, Includes the multi-mode peak detection device for SPAD detectors as described in claim 9.

Images