A bioacoustic complexity index method with low sensitivity to noise

Abstract

This invention discloses a bioacoustic complexity index method with low sensitivity to noise, comprising: performing a short-time discrete Fourier transform on the field bioacoustic monitoring recording data during the analysis period to obtain its time-frequency power spectrum; estimating the narrowband noise average power of each frequency point using a set of noise time-frequency points without bioacoustic signals; performing power spectrum subtraction on each frequency point to obtain the noise-enhanced bioacoustic time-frequency power spectrum; subdividing the analysis bandwidth and analysis period into time-frequency domains according to a certain frequency step and time step, and calculating the time-frequency partition index value; and summing all the time-frequency partition index values ​​within the analysis bandwidth and analysis period to obtain the final bioacoustic index value, i.e., the noise-reduced bioacoustic complexity index based on the change in bioacoustic time-frequency power after noise enhancement. This invention significantly reduces the sensitivity of the index results to noise and significantly expands the spatiotemporal application range of similar acoustic index strategies.

Description

Technical Field

[0001] This invention relates to the fields of rapid assessment of biological species diversity and acoustic signal processing, specifically to a bioacoustic complexity index method that is insensitive to noise. Background Technology

[0002] Many organisms in nature emit sounds, which are rich in important biological information. Compared with traditional field survey methods, Passive Acoustic Monitoring (PAM) has advantages such as low overall cost, minimal habitat disturbance, and the ability to provide long-term data recording, and has enormous potential for development in biodiversity conservation, natural resource management, and ecosystem monitoring. With the increasing prevalence of PAM, the need for efficient analysis and processing of field monitoring data to quickly extract meaningful biological and ecological information from natural soundscapes is growing stronger. The acoustic index method, which has emerged in recent years, is precisely such a technique that can achieve rapid biodiversity assessment by specifically measuring and quantifying the characteristics of biological acoustic signals in actual recordings, and it is receiving increasing attention and application both domestically and internationally.

[0003] In the field of ecological monitoring, more than 60 bioacoustic indices have been developed. Among them, the Acoustic Complexity Index (ACI) (Pieretti N, Farina A, Morri DA new methodology to infer the singing activity of an avian community: The Acoustic Complexity Index (ACI)[J]. Ecological indicators, 2011, 11(3):868-873.) is one of the most studied, researched, and widely used indices. This index is often considered to be able to capture the dynamic changes and complexity of biological communities and ecosystems, and is often used to quickly assess the species richness of habitats. However, in practical applications, whether there is a clear relationship between ACI and species richness, and its overall effectiveness in different ecosystems and biological communities, has always been highly controversial. This is because the field monitoring recording data actually used to calculate ACI is not a pure bioacoustic signal, but a mixed signal containing both bioacoustic sound and various noise components. Although the ACI (Audio-Induced Contrast) incorporates some considerations for noise reduction and line spectrum interference suppression in its design, its exponential calculation process, based on time-frequency intensity changes rather than power changes, inherently limits its noise suppression capabilities. This makes the ACI highly sensitive to noise interference and its variations. Its value deviates further from the ideal value corresponding to a pure biological sound signal under noise-free conditions as the signal-to-noise ratio (SNR) in the actual soundscape decreases. Since the intensity of actual biological sound signals is affected by various uncontrollable factors such as the volume, distance, and number of sound sources, and the soundscape environment, weather conditions, and biological populations vary greatly at different monitoring locations and time periods, the biological sound SNR in actual field recordings fluctuates significantly and randomly over time and space. Furthermore, the biological sound SNR is relatively low for most of the data period, rarely exceeding 20 dB. Therefore, the high sensitivity of the ACI value to changes in biological sound SNR makes it unsuitable for stable and reliable biodiversity monitoring, hindering its large-scale application in the field of ecological monitoring. Therefore, in order to enhance the universality of ACI's index design strategy in various ecological monitoring soundscapes and make it a truly reliable and effective tool for rapid biodiversity assessment, it is urgent to improve its processing to enhance its robustness to noise impact. Summary of the Invention

[0004] The purpose of this invention is to provide a bioacoustic complexity index method that is less sensitive to noise.

[0005] The technical solution to achieve the objective of this invention is: a bioacoustic complexity index method with low sensitivity to noise, comprising the following steps:

[0006] Step 1: Perform short-time discrete Fourier transform on the acoustic recording data of field organisms during the analysis period to obtain its time-frequency power spectrum;

[0007] Step 2: Using the time-frequency power spectrum of the recording data obtained in Step 1, and using the set of noise time-frequency points that do not contain biological sound signals, estimate the average power of the narrowband noise at each frequency point.

[0008] Step 3: Using the time-frequency power spectrum of the recording data obtained in Step 1 and the average power of narrowband noise at each frequency point obtained in Step 2, perform power spectrum subtraction on each frequency point to obtain the time-frequency power spectrum of the bio-sound after noise reduction and enhancement.

[0009] Step 4: Using the noise-reduced and enhanced bio-sound time-frequency power spectrum obtained in Step 3, the analysis bandwidth and analysis period are subdivided into time-frequency domains according to a certain frequency step and time step. Each time-frequency partition corresponds to several frequency points and several time frames. The sum of the absolute values ​​of the power differences between all adjacent time-frequency points in each time-frequency partition is used as the numerator, and the sum of the power of all time-frequency points is used as the denominator. The time-frequency partition index value is obtained by calculating the ratio between the two.

[0010] Step 5: Using all the time-frequency partition index values ​​obtained in Step 4, sum up the analysis bandwidth and all the time-frequency partition index values ​​within the analysis period to obtain the final bioacoustic index value, which is the noise reduction complexity index based on the change in bioacoustic time-frequency power after noise reduction enhancement.

[0011] Furthermore, the specific process of step 1 is as follows:

[0012] The field biological acoustic monitoring recording data during the analysis period were processed by frame segmentation. Each frame of data was windowed and then subjected to short-time discrete Fourier transform to obtain the time-frequency power spectrum P(k,q), 1≤k≤K, 1≤q≤Q, where k and q are the frame number and frequency point number, respectively, K is the total number of frames in the analysis period, and Q is the total number of frequency points in the analysis bandwidth.

[0013] Furthermore, the specific process of step 2 is as follows:

[0014] Step 2-1: Obtain the set of background noise time-frequency points based on the time-frequency power spectrum of the recording data. The set of pure noise time-frequency points without biological sound in the q-th frequency point is denoted as Ω(q), and the number of noise time-frequency points in the set Ω(q) is denoted as N(q).

[0015] Step 2-2: Based on the set of noise time-frequency point frame numbers Ω(q) and the number of noise time-frequency points N(q) obtained in Step 2-1, calculate the average narrowband noise power at each frequency point:

[0016]

[0017] Furthermore, the specific process of step 3 is as follows:

[0018] Using the time-frequency power spectrum of the bio-sound monitoring recording data obtained in step 1 and the average power of narrowband noise at each frequency point obtained in step 2, a relaxation factor γ not greater than 1 is set, and then the time-frequency power spectrum P′(k,q) of the bio-sound after noise reduction and enhancement is calculated using the spectral subtraction method, with zero as the lower limit of the time-frequency power after noise reduction:

[0019]

[0020] Furthermore, the specific process of step 4 is as follows:

[0021] Step 4-1: Subdivide the analysis bandwidth and analysis period in the time and frequency domain according to the set frequency step size and time step size. Each frequency step size includes i frequency points, and the set of frequency points in the nth frequency step size is:

[0022] Ω n ={q|(n-1)i<q≤ni}

[0023] Each time step includes j time frames, then the set of time frames in the m-th time step is:

[0024] Ψ m ={k|(m-1)j<k≤mj}

[0025] Step 4-2: Calculate the sum of the absolute values ​​of the power differences between all adjacent time-frequency points in each time-frequency partition, D(m,n), and the total power A(m,n) of all time-frequency points:

[0026]

[0027] Step 4-3: Calculate the ratio of the sum of absolute power differences D(m,n) in each time-frequency zone to the total power sum A(m,n) to obtain the time-frequency zone index value. If the total power sum is 0, the time-frequency zone is considered a completely noisy zone with no biological sound activity, and the index value is directly set to 0.

[0028]

[0029] Furthermore, the specific process of step 5 is as follows:

[0030] The final bioacoustic index value (DACI) is obtained by summing the analysis bandwidth and all time-frequency partition index values ​​within the analysis period.

[0031]

[0032] In the formula, n start n end These are the starting and ending frequency step numbers for the analysis bandwidth, respectively, m start m end These are the start and end time step numbers for the analysis period, respectively.

[0033] An electronic device includes a microphone, a memory, a processor, and a computer program stored in the memory and executable on the processor, wherein the processor, when executing the program, implements the above-described method for a bioacoustic complexity index that is insensitive to noise.

[0034] A computer-readable storage medium having a computer program stored thereon that, when executed by a processor, implements the above-described method for a bioacoustic complexity index that is insensitive to noise.

[0035] A computer program product includes a computer program that, when executed by a processor, implements the aforementioned method for a bioacoustic complexity index that is insensitive to noise.

[0036] Compared with existing technologies, the beneficial effects of this invention are as follows: 1) It corrects the design defects of conventional ACI in principle and proposes a new strategy for bioacoustic complexity index with low sensitivity to noise, namely, the noise reduction complexity index DCI based on the time-frequency power change of bioacoustic sound after noise reduction and enhancement; 2) Without affecting the existing ecological interpretation and research accumulation of conventional ACI, it significantly improves the robustness of the actual index value to the large-scale random changes of bioacoustic SNR over time, which helps to realize large-scale promotion and application in the field of ecological monitoring; 3) It is simple to calculate and easy to implement, providing a reliable and robust solution for efficient analysis and processing of large-scale field monitoring recording data.

[0037] The present invention will now be described in further detail with reference to the accompanying drawings. Attached Figure Description

[0038] Figure 1 This is a flowchart of the method of the present invention.

[0039] Figure 2 The time-domain waveforms and corresponding time-spectrum diagrams of synthesized noisy bird sound signals obtained by superimposing clean bird sound signal data with the same time-frequency distribution structure onto the same noise background using two preset SNRs of 40dB and -20dB respectively.

[0040] Figure 3 This is a graph showing the change of the exponential values ​​calculated by ACI and DAI for synthesized noisy bird sound signal data with the same time-frequency distribution as a function of the bird sound SNR.

[0041] Figure 4 When the SNR of bird calls is set to 40dB and -20dB respectively, the curves of the time-frequency partition index values ​​calculated by ACI and DAI for synthesized noisy bird call signal data with the same time-frequency distribution of bird call signals are shown along the frequency axis.

[0042] Figure 5 This is a comparison chart of the distribution of ACI and DCI values ​​calculated using 200 sets of randomly selected actual field monitoring recordings with a high SNR of bird calls above 25dB. Detailed Implementation

[0043] This invention addresses the needs of ecological acoustic monitoring and quality assessment by proposing a bioacoustic complexity index with low sensitivity to noise: the Denoising Acoustic Complexity Index (DACI), which is based on the change in time-frequency power of bioacoustic sounds after noise reduction and enhancement. DCI utilizes the statistical independence between bioacoustic and noise signals and the statistical property that the power of noisy bioacoustic sounds equals the sum of the bioacoustic and noise power. By replacing the intensity subtraction operation of adjacent time-frequency points in ACI with a power subtraction operation after spectral subtraction and noise reduction, it effectively suppresses the influence of background noise on the index results while preserving as much of the time-frequency structure information and inherent power changes of bioacoustic sounds as possible. This allows for the acquisition of acoustic statistical characteristics of biological communities approaching ideal noise-free conditions. The proposed DCI exhibits high numerical robustness even at bioacoustic SNRs as low as -20 dB and shows a strong correlation with the results of conventional ACI under high SNR conditions. Therefore, this invention can significantly improve the spatiotemporal universality of similar bioacoustic index design routes in various ecological monitoring acoustic environments without affecting the existing ecological interpretation and research accumulation of conventional ACI, providing an efficient and feasible new approach and a reliable and robust new tool for the large-scale promotion of rapid biodiversity assessment in the field of ecological monitoring.

[0044] Combination Figure 1 This invention is a bioacoustic complexity index method with low sensitivity to noise. Given that birds are highly sensitive indicator organisms for ecological monitoring, the following uses birds as the research object and illustrates the specific implementation of this invention through a specific example of rapid assessment of bird song diversity in the wild. The steps are as follows:

[0045] Step 1: Perform a short-time discrete Fourier transform on the recorded bird calls during the analysis period to obtain its time-frequency power spectrum:

[0046] The recorded data of bird calls in the field during the analysis period are processed by frame segmentation. Each frame of data is windowed and then subjected to short-time discrete Fourier transform to obtain the time-frequency power spectrum P(k,q), 1≤k≤K, 1≤q≤Q, where k and q are the frame number and frequency point number, respectively, K is the total number of frames in the analysis period, and Q is the total number of frequency points in the analysis bandwidth. In a specific example of this invention, the data sampling rate is 32kHz, the analysis period length is 60s, the frame length and frame shift are both 16ms, and the corresponding K and Q are 3750 and 256, respectively. The number of samples for the short-time discrete Fourier transform is 512, and a Hamming window is added before the short-time discrete Fourier transform.

[0047] Step 2: Using the time-frequency power spectrum of the recording data obtained in Step 1, and using the set of noise time-frequency points that do not contain bird sounds, estimate the average narrowband noise power at each frequency point:

[0048] Step 2-1: Use the energy probability histogram to set dual thresholds for background noise extraction and obtain a set of noise time-frequency points that do not contain bird sounds. The specific implementation method is as follows: perform probability statistics on the power of all frames corresponding to each frequency point, find the power T corresponding to the maximum probability value from the probability histogram, set a high threshold T1 = T + 0.1 and a low threshold T2 = T + 0.001, mark the time-frequency points with power lower than the low threshold T2 as noise time-frequency points, and form a set of the frame numbers of the noise time-frequency points marked by each frequency point. The set of the frame numbers of the noise time-frequency points corresponding to the qth frequency point is denoted as Ω(q), and the number of noise time-frequency points in the set Ω(q) is denoted as N(q).

[0049] Step 2-2: Based on the set of noise time-frequency point frame numbers Ω(q) and the number of noise time-frequency points N(q) obtained in Step 2-1, calculate the average narrowband noise power at each frequency point:

[0050]

[0051] Step 3: Using the time-frequency power spectrum of the recording data obtained in Step 1 and the average power of narrowband noise at each frequency point obtained in Step 2, perform power spectrum subtraction on each frequency point to obtain the noise-reduced and enhanced bird sound time-frequency power spectrum:

[0052] Using the time-frequency power spectrum of the bird call monitoring data obtained in step 1 and the average power of the narrowband noise obtained in step 2, the relaxation factor γ is set to 0.618 in this specific example of the invention. Then, the spectral subtraction method is used to calculate the noise-reduced and enhanced bird call time-frequency power spectrum, with zero as the lower limit of the noise-reduced time-frequency power.

[0053]

[0054] Step 4: Using the noise-reduced and enhanced bird sound time-frequency power spectrum obtained in Step 3, the analysis bandwidth and analysis period are subdivided into time-frequency domains according to a certain frequency step size and time step size. Each time-frequency partition corresponds to several frequency points and several time frames. The sum of the absolute values ​​of the power differences between all adjacent time-frequency points in each time-frequency partition is used as the numerator, and the sum of the power of all time-frequency points is used as the denominator. The time-frequency partition index value is obtained by calculating the ratio of the two.

[0055] Step 4-1: Subdivide the analysis bandwidth and analysis period in the time and frequency domain according to a certain frequency step and time step. Each frequency step includes i frequency points. In the specific example of this invention, i is 1, corresponding to a frequency step of 62.5Hz. Then, the set of frequency points in each frequency step contains only 1 frequency point element. The set of frequency points in the nth frequency step is:

[0056] Ω n ={q|(n-1)i<q≤ni}={q|q=n}

[0057] Each time step includes j time frames. In a specific embodiment of this invention, j is 313, corresponding to a time step of approximately 5 seconds. Therefore, each time step contains 313 time frame elements. The set of time frames in the m-th time step is as follows:

[0058] Ψ m ={k|(m-1)j<k≤mj}

[0059] Step 4-2: Calculate the sum of the absolute values ​​of the power differences between all adjacent time-frequency points in each time-frequency partition, D(m,n), and the total power A(m,n) of all time-frequency points:

[0060]

[0061] Step 4-3: Calculate the ratio of the sum of absolute power differences D(m,n) in each time-frequency zone to the total power sum A(m,n) to obtain the time-frequency zone index value. If the total power sum is 0, the time-frequency zone is considered a completely noisy zone with no bird activity, and the index value is directly set to 0.

[0062]

[0063] Step 5: Using all the time-frequency partition index values ​​obtained in Step 4, sum the analysis bandwidth and all the time-frequency partition index values ​​within the analysis period to obtain the final bioacoustic index value, i.e., the noise reduction acoustic complexity index (DACI) based on the time-frequency power change of bioacoustic sound after noise reduction enhancement.

[0064]

[0065] In the formula, n start nend These are the starting and ending frequency step numbers for the analysis bandwidth, respectively, m start m end These are the start and end time step numbers for the analysis period, respectively. In a specific embodiment of this invention, the analysis bandwidth is 500Hz-12kHz, corresponding to n. start Take 9, n end Take 192, the analysis period is 60s, and the corresponding m start Take 1, m end Take 12.

[0066] In one embodiment, an electronic device is provided, including a microphone, a memory, a processor, and a computer program stored in the memory and executable on the processor, wherein the processor, when executing the program, implements the above-described method for a bioacoustic complexity index that is insensitive to noise.

[0067] In one embodiment, a computer-readable storage medium is provided having a computer program stored thereon that, when executed by a processor, implements the above-described method for a bioacoustic complexity index that is insensitive to noise.

[0068] In one embodiment, a computer program product is provided, including a computer program that, when executed by a processor, implements the above-described method for a bioacoustic complexity index that is insensitive to noise.

[0069] Figure 2 In the diagram, (a) and (b) are the time-domain waveforms and corresponding time-spectrum diagrams of synthesized noisy bird sound signals obtained by superimposing clean bird sound signal data with the same time-frequency distribution structure onto the same noise background using two preset SNR values ​​of 40dB and -20dB, respectively. Under the high SNR condition of 40dB, the bird sound signal is clearly visible, while under the low SNR condition of -20dB, the bird sound signal is almost completely masked by the background noise.

[0070] Figure 3 The graphs of the exponent values ​​calculated by ACI and DAI for synthesized noisy bird call signal data with the same time-frequency distribution are presented as a function of the bird call signal SNR. It can be seen that when the bird call signals have the same time-frequency distribution, ACI shows a significant monotonically decreasing change with decreasing SNR, while the calculation result of DAI shows no significant change even when the SNR is as low as -20dB. This indicates that ACI is more sensitive to noise, while DAI exhibits significantly better noise robustness than ACI.

[0071] Figure 4The graphs show the variation of the time-frequency zoning index values ​​of ACI and DAI along the frequency axis for synthesized noisy bird sound signal data with the same time-frequency distribution, with bird sound SNR set to 40dB and -20dB respectively. It can be seen that under the low SNR condition, the ACI time-frequency zoning index values ​​for all frequency bands are lower than those under the high SNR condition, while the DAI time-frequency zoning indices maintain a high degree of consistency in both magnitude and trend along the frequency axis under both high and low SNR conditions, with the two curves showing a very high degree of fit. This figure demonstrates that DAI not only exhibits high robustness in the overall numerical results across the entire frequency band but also shows high consistency in results for each individual frequency component.

[0072] Figure 5 A comparative chart of the numerical distributions of ACI and DCI, calculated using 200 randomly selected sets of actual field monitoring recordings with high SNR (Sound Frequency Ratio) of over 25 dB, is presented. It can be seen that under high SNR conditions, the numerical trends of DCI and ACI are basically consistent, and the calculated Spearman correlation coefficient between the two is 0.83, demonstrating a strong correlation. This result indicates that although DCI focuses on time-frequency power variations rather than intensity variations, its technical approach is rooted in the theoretical assumptions and design concepts of ACI. Therefore, the introduction of DCI does not affect the existing ecological interpretations and research accumulation of ACI.

Claims

1. A bioacoustic complexity index method with low sensitivity to noise effects, characterized in that, Includes the following steps: Step 1: Perform short-time discrete Fourier transform on the acoustic recording data of field organisms during the analysis period to obtain its time-frequency power spectrum; Step 2: Using the time-frequency power spectrum of the recording data obtained in Step 1, and using the set of noise time-frequency points that do not contain biological sound signals, estimate the average power of the narrowband noise at each frequency point. Step 3: Using the time-frequency power spectrum of the recording data obtained in Step 1 and the average power of narrowband noise at each frequency point obtained in Step 2, perform power spectrum subtraction on each frequency point to obtain the time-frequency power spectrum of the bio-sound after noise reduction and enhancement. Step 4: Using the noise-reduced and enhanced bio-sound time-frequency power spectrum obtained in Step 3, the analysis bandwidth and analysis period are subdivided into time-frequency domains according to the set frequency step and time step. Each time-frequency partition corresponds to several frequency points and several time frames. The sum of the absolute values ​​of the power differences between all adjacent time-frequency points in each time-frequency partition is used as the numerator, and the sum of the power of all time-frequency points is used as the denominator. The time-frequency partition index value is obtained by calculating the ratio of the two. Step 5: Using all the time-frequency partition index values ​​obtained in Step 4, sum up the analysis bandwidth and all the time-frequency partition index values ​​within the analysis period to obtain the final bioacoustic index value, which is the noise reduction complexity index based on the change in bioacoustic time-frequency power after noise reduction enhancement.

2. The bioacoustic complexity index method with low sensitivity to noise effects according to claim 1, characterized in that: The specific process of step 1 is as follows: The field biological acoustic monitoring recording data during the analysis period were processed by frame segmentation. Each frame of data was windowed and then subjected to short-time discrete Fourier transform to obtain the time-frequency power spectrum P(k,q), 1≤k≤K, 1≤q≤Q, where k and q are the frame number and frequency point number, respectively, K is the total number of frames in the analysis period, and Q is the total number of frequency points in the analysis bandwidth.

3. The bioacoustic complexity index method with low sensitivity to noise effects according to claim 2, characterized in that: The specific process of step 2 is as follows: Step 2-1: Obtain the set of background noise time-frequency points based on the time-frequency power spectrum of the recording data. The set of pure noise time-frequency points without biological sound in the q-th frequency point is denoted as Ω(q), and the number of noise time-frequency points in the set Ω(q) is denoted as N(q). Step 2-2: Based on the set of noise time-frequency point frame numbers Ω(q) and the number of noise time-frequency points N(q) obtained in Step 2-1, calculate the average narrowband noise power at each frequency point:

4. The bioacoustic complexity index method with low sensitivity to noise effects according to claim 3, characterized in that: The specific process of step 3 is as follows: Using the time-frequency power spectrum of the bio-sound monitoring recording data obtained in step 1 and the average power of narrowband noise at each frequency point obtained in step 2, a relaxation factor γ not greater than 1 is set, and then the time-frequency power spectrum P′(k,q) of the bio-sound after noise reduction and enhancement is calculated using the spectral subtraction method, with zero as the lower limit of the time-frequency power after noise reduction:

5. The bioacoustic complexity index method with low sensitivity to noise effects according to claim 4, characterized in that; The specific process of step 4 is as follows: Step 4-1: Subdivide the analysis bandwidth and analysis period in the time and frequency domain according to the set frequency step size and time step size. Each frequency step size includes i frequency points, and the set of frequency points in the nth frequency step size is: Ω n ={q|(n-1)i<q≤ni} Each time step includes j time frames, then the set of time frames in the m-th time step is: P m ={k|(m-1)j<k≤mj} Step 4-2: Calculate the sum of the absolute values ​​of the power differences between all adjacent time-frequency points in each time-frequency partition, D(m,n), and the total power A(m,n) of all time-frequency points: Step 4-3: Calculate the ratio of the sum of absolute power differences D(m,n) in each time-frequency zone to the total power sum A(m,n) to obtain the time-frequency zone index value. If the total power sum is 0, the time-frequency zone is considered a completely noisy zone with no biological sound activity, and the index value is directly set to 0.

6. The bioacoustic complexity index method with low sensitivity to noise effects according to claim 5, characterized in that; The specific process of step 5 is as follows: The final bioacoustic index value (DACI) is obtained by summing the analysis bandwidth and all time-frequency partition index values ​​within the analysis period. In the formula, n start n end These are the starting and ending frequency step numbers for the analysis bandwidth, respectively, m start m end These are the start and end time step numbers for the analysis period, respectively.

7. An electronic device comprising a microphone, a memory, a processor, and a computer program executable on the processor, characterized in that, When the processor executes the program, it implements the steps of the method as described in any one of claims 1-6.

8. A computer-readable storage medium having a computer program stored thereon, characterized in that, When the program is executed in a processor, it implements the steps of the method as described in any one of claims 1-6.

9. A computer program product, comprising a computer program, characterized in that, When executed by a processor, the computer program implements the steps of the method described in any one of claims 1-6.

Images