A music database retrieval method and system based on feature extraction

Abstract

The present application discloses a music database retrieval method and system based on feature extraction, which relates to the technical field of retrieval analysis, and includes obtaining feature data corresponding to a target music file, the feature data including spectrum data, rhythm data and harmony data; collecting and analyzing the spectrum data corresponding to the target music file, and confirming the spectrum evaluation coefficient corresponding to the target music file; collecting and analyzing the rhythm data corresponding to the target music file, and confirming the rhythm evaluation coefficient corresponding to the target music file; collecting and analyzing the harmony data corresponding to the target music file, and confirming the harmony evaluation coefficient corresponding to the target music file; confirming the comprehensive evaluation coefficient corresponding to the target music file, and searching a preset music database according to the comprehensive evaluation coefficient corresponding to the target music file, and then matching and confirming the corresponding target music file. The present application has the effect of improving the accuracy of retrieval.

Description

Technical Field

[0001] The present application relates to the technical field of retrieval analysis, and in particular to a music database retrieval method and system based on feature extraction. Background Art

[0002] Database technology is a technology that studies the basic theories and implementation methods of database structure, storage, design, management and application, and uses these theories to process, analyze and understand the data in the database. That is, database technology is a software science that studies, manages and applies databases. The object of database technology research and management is data, so the specific content involved in database technology mainly includes: through unified organization and management of data, establish corresponding databases and data warehouses according to the specified structure; use database management systems and data mining systems to design data management and data mining application systems that can realize multiple functions such as adding, modifying, deleting, processing, analyzing, understanding, reporting and printing data in the database; and use application management systems to finally realize data processing, analysis and understanding.

[0003] In recent years, there are more and more ways to spread music through the Internet. However, due to the particularity of music information, in most cases people can only generate specific feelings about it after hearing it. Therefore, there is still a lack of music database systems that can be searched for music use in specific situations.

[0004] In the related art, when searching the music database, the music features are not extracted and analyzed from multiple aspects, resulting in reduced accuracy in music judgment, which in turn reduces the precision of the search, and there is room for improvement. Summary of the invention

[0005] In view of the deficiencies in the prior art, the present application provides a music database retrieval method and system based on feature extraction.

[0006] In a first aspect, the present application provides a music database retrieval method based on feature extraction, comprising the following steps:

[0007] Acquire feature data corresponding to the target music file, wherein the feature data includes spectrum data, rhythm data and harmony data;

[0008] Collecting and analyzing the spectrum data corresponding to the target music file, and then confirming the spectrum evaluation coefficient corresponding to the target music file;

[0009] Collecting and analyzing the rhythm data corresponding to the target music file, and then confirming the rhythm evaluation coefficient corresponding to the target music file;

[0010] Collecting and analyzing harmony data corresponding to the target music file, and then determining a harmony evaluation coefficient corresponding to the target music file;

[0011] A comprehensive analysis is performed on the spectrum evaluation coefficient, rhythm evaluation coefficient and harmony evaluation coefficient corresponding to the target music file to confirm the comprehensive evaluation coefficient corresponding to the target music file, and a preset music database is searched based on the comprehensive evaluation coefficient corresponding to the target music file to match and confirm the corresponding target music file.

[0012] Preferably, the spectrum data includes spectrum, spectrum centroid data and spectrum bandwidth data, the rhythm data includes beat number, beat speed and beat intensity, and the harmony data includes pitch data.

[0013] Preferably, collecting and analyzing the spectrum data corresponding to the target music file, and then confirming the spectrum evaluation coefficient corresponding to the target music file, specifically includes:

[0014] Extracting frequency component data from the frequency spectrum corresponding to the frequency spectrum data, and normalizing the frequency component data, wherein the frequency component data includes the number of frequency components, the frequency and amplitude corresponding to each frequency component;

[0015] By calculating the formula , get the spectrum centroid data corresponding to the target music file ,in, It is represented by the number corresponding to each frequency component. , Respectively expressed as The frequency and amplitude corresponding to each frequency component;

[0016] By calculating the formula , get the spectrum bandwidth data corresponding to the target music file ;

[0017] The spectral centroid data corresponding to the target music file Spectral bandwidth data corresponding to the target music file Substitute into the formula , and obtain the spectrum evaluation coefficient corresponding to the target music file ,in, Respectively represent the preset reference spectrum centroid data and reference spectrum bandwidth data, They are respectively expressed as the weight coefficients corresponding to the spectrum centroid data and the spectrum bandwidth data.

[0018] Preferably, collecting and analyzing the rhythm data corresponding to the target music file, and then confirming the rhythm evaluation coefficient corresponding to the target music file, specifically includes:

[0019] Extract the number of beats corresponding to the target music file from the rhythm data corresponding to the target music file , , and then through the formula Determine the duration of each beat ,in, Respectively expressed as The start time corresponding to each beat;

[0020] By formula , get the tempo of the target music file , Expressed as The beat speed correction coefficient corresponding to each beat;

[0021] Obtain the audio signal corresponding to each beat, and confirm the beat intensity corresponding to each beat based on the audio signal corresponding to each beat and the short-time energy method ;

[0022] By calculating the formula , and obtain the rhythm evaluation coefficient corresponding to the target music file ,in, They represent the preset reference tempo speed and reference tempo intensity respectively. They are respectively expressed as the weight coefficients corresponding to the beat speed and beat intensity.

[0023] Preferably, collecting and analyzing the harmony data corresponding to the target music file, and then confirming the harmony evaluation coefficient corresponding to the target music file, specifically includes:

[0024] In the preset time window, the pitch data corresponding to the target music file is collected in real time, and then the pitch change curve over time is confirmed. ;

[0025] By formula , and obtain the harmony evaluation coefficient corresponding to the target music file ,in, Represents a preset reference pitch change curve.

[0026] Preferably, confirming the comprehensive evaluation coefficient corresponding to the target music file specifically includes:

[0027] The spectral evaluation coefficient corresponding to the target music file , the rhythm evaluation coefficient corresponding to the target music file Harmony evaluation coefficient corresponding to the target music file Substitute into the formula , and obtain the comprehensive evaluation coefficient corresponding to the target music file ,in, They are respectively expressed as weight coefficients corresponding to the spectrum evaluation coefficient, rhythm evaluation coefficient, and harmony evaluation coefficient.

[0028] Preferably, searching a preset music database according to the comprehensive evaluation coefficient corresponding to the target music file, and then matching and confirming the corresponding target music file, specifically includes:

[0029] The comprehensive evaluation coefficient corresponding to the target music file Compare with the preset comprehensive evaluation interval to confirm the comprehensive evaluation coefficient corresponding to the target music file The target comprehensive assessment range;

[0030] Obtaining the comprehensive evaluation coefficient corresponding to each music file in the music database, screening out the music files whose comprehensive evaluation coefficients are within the target comprehensive evaluation interval, and storing the music files in the music file set;

[0031] Obtaining the number of elements in the music file set, and if the number of elements in the music file set is equal to a preset number threshold, determining that the music file in the music file set is the corresponding target music file;

[0032] If the number of elements in the music file set exceeds a preset number threshold, each element in the music file set is judged based on the first search method, and the corresponding target music file is matched and confirmed based on the judgment result.

[0033] In a second aspect, the present application provides a music database retrieval system based on feature extraction, comprising:

[0034] A feature acquisition module, used to acquire feature data corresponding to a target music file, wherein the feature data includes spectrum data, rhythm data and harmony data;

[0035] A spectrum data analysis module, used to collect and analyze the spectrum data corresponding to the target music file, and then determine the spectrum evaluation coefficient corresponding to the target music file;

[0036] A rhythm data analysis module, used to collect and analyze the rhythm data corresponding to the target music file, and then determine the rhythm evaluation coefficient corresponding to the target music file;

[0037] A harmony data analysis module, used to collect and analyze the harmony data corresponding to the target music file, and then determine the harmony evaluation coefficient corresponding to the target music file;

[0038] The retrieval and matching module is used to comprehensively analyze the spectral evaluation coefficient, rhythm evaluation coefficient and harmony evaluation coefficient corresponding to the target music file, confirm the comprehensive evaluation coefficient corresponding to the target music file, and search the preset music database according to the comprehensive evaluation coefficient corresponding to the target music file, and then match and confirm the corresponding target music file.

[0039] In a third aspect, the present application provides a computer-readable storage medium storing instructions, which, when executed on a computer, enable the computer to execute any one of the above-described feature extraction-based music database retrieval methods.

[0040] In summary, the present application includes at least one of the following beneficial technical effects:

[0041] 1. The present invention provides a music database retrieval method based on feature extraction, which obtains the spectrum data, rhythm data and harmony data corresponding to the target music file, collects and analyzes the spectrum data, rhythm data and harmony data corresponding to the target music file, confirms the spectrum evaluation coefficient, rhythm evaluation coefficient and harmony evaluation coefficient corresponding to the target music file, and then confirms the comprehensive evaluation coefficient corresponding to the target music file, searches a preset music database through the comprehensive evaluation coefficient corresponding to the target music file, and then matches and confirms the corresponding target music file, thereby effectively improving the accuracy of the retrieval;

[0042] 2. By screening out music files whose comprehensive evaluation coefficients are within the target comprehensive evaluation range and storing the music files in a music file set, the number of elements in the music file set is further confirmed, and the corresponding target music files are judged based on the number of elements in the music file set, thereby effectively improving the accuracy of the retrieval. BRIEF DESCRIPTION OF THE DRAWINGS

[0043] Figure 1 This is a flow chart of the method for music database retrieval based on feature extraction in an embodiment of the present application.

[0044] Figure 2 It is a system diagram of music database retrieval based on feature extraction in an embodiment of the present application. DETAILED DESCRIPTION

[0045] The following is combined with Figure 1-2 This application is described in further detail.

[0046] Example 1

[0047] The embodiment of the present application discloses a music database retrieval method based on feature extraction.

[0048] Reference Figure 1, a music database retrieval method based on feature extraction, comprising the following steps:

[0049] S1. Obtaining characteristic data corresponding to the target music file, wherein the characteristic data includes spectrum data, rhythm data and harmony data;

[0050] S2. Collect and analyze the spectrum data corresponding to the target music file, and then confirm the spectrum evaluation coefficient corresponding to the target music file;

[0051] S3. Collect and analyze the rhythm data corresponding to the target music file, and then confirm the rhythm evaluation coefficient corresponding to the target music file;

[0052] S4. Collect and analyze the harmony data corresponding to the target music file, and then confirm the harmony evaluation coefficient corresponding to the target music file;

[0053] S5. Comprehensively analyze the spectrum evaluation coefficient, rhythm evaluation coefficient and harmony evaluation coefficient corresponding to the target music file, confirm the comprehensive evaluation coefficient corresponding to the target music file, and search the preset music database according to the comprehensive evaluation coefficient corresponding to the target music file, and then match and confirm the corresponding target music file.

[0054] Through the above technical scheme, by obtaining the spectral data, rhythm data and harmony data corresponding to the target music file, and collecting and analyzing the spectral data, rhythm data and harmony data corresponding to the target music file, the spectral evaluation coefficient, rhythm evaluation coefficient and harmony evaluation coefficient corresponding to the target music file are confirmed, and then the comprehensive evaluation coefficient corresponding to the target music file is confirmed. The preset music database is searched through the comprehensive evaluation coefficient corresponding to the target music file, and then the corresponding target music file is matched and confirmed, thereby effectively improving the accuracy of the retrieval.

[0055] Furthermore, the spectrum data includes spectrum, spectrum centroid data and spectrum bandwidth data, the rhythm data includes beat number, beat speed and beat intensity, and the harmony data includes pitch data.

[0056] Specifically, the characteristic data corresponding to the target music file includes but is not limited to spectrum data, rhythm data and harmony data. The spectrum data, rhythm data and harmony data are analyzed to determine the comprehensive evaluation coefficient corresponding to the target music file.

[0057] It should be noted that collecting and analyzing the spectrum data corresponding to the target music file, and then confirming the spectrum evaluation coefficient corresponding to the target music file, specifically includes:

[0058] Extracting frequency component data from the frequency spectrum corresponding to the frequency spectrum data, and normalizing the frequency component data, wherein the frequency component data includes the number of frequency components, the frequency and amplitude corresponding to each frequency component;

[0059] Specifically, frequency component data is extracted from the frequency spectrum corresponding to the frequency spectrum data by using Fourier transform technology.

[0060] By calculating the formula , get the spectrum centroid data corresponding to the target music file ,in, It is represented by the number corresponding to each frequency component. , Respectively expressed as The frequency and amplitude corresponding to each frequency component;

[0061] By calculating the formula , get the spectrum bandwidth data corresponding to the target music file ;

[0062] The spectral centroid data corresponding to the target music file Spectral bandwidth data corresponding to the target music file Substitute into the formula , and obtain the spectrum evaluation coefficient corresponding to the target music file ,in, Respectively represent the preset reference spectrum centroid data and reference spectrum bandwidth data, They are respectively expressed as the weight coefficients corresponding to the spectrum centroid data and the spectrum bandwidth data.

[0063] Specifically, spectrum refers to the representation of audio signals in the frequency domain, including the frequencies and amplitudes corresponding to different frequency components; the spectrum centroid is represented by the centroid position corresponding to the spectrum energy, that is, a high spectrum centroid indicates that the audio signal has more high-frequency components, and a low spectrum centroid indicates that the audio signal has more low-frequency components; the spectrum bandwidth is represented by the width of the spectrum energy distribution, that is, the wider the spectrum width, the more complex the timbre.

[0064] Specifically, by confirming the frequency and amplitude corresponding to each frequency component, the spectrum centroid data corresponding to the target music file and the spectrum bandwidth data corresponding to the target music file are confirmed, thereby effectively improving the authenticity of the data and further confirming the spectrum evaluation coefficient corresponding to the target music file.

[0065] It should be noted that collecting and analyzing the rhythm data corresponding to the target music file, and then confirming the rhythm evaluation coefficient corresponding to the target music file, specifically includes:

[0066] Extract the number of beats corresponding to the target music file from the rhythm data corresponding to the target music file , , and then through the formula Determine the duration of each beat ,in, Respectively expressed as The start time corresponding to each beat;

[0067] By formula , get the tempo of the target music file , Expressed as The beat speed correction coefficient corresponding to each beat;

[0068] Obtain the audio signal corresponding to each beat, and confirm the beat intensity corresponding to each beat based on the audio signal corresponding to each beat and the short-time energy method ;

[0069] By calculating the formula , and obtain the rhythm evaluation coefficient corresponding to the target music file ,in, They represent the preset reference tempo speed and reference tempo intensity respectively. They are respectively expressed as the weight coefficients corresponding to the beat speed and beat intensity.

[0070] Furthermore, collecting and analyzing the harmony data corresponding to the target music file, and then confirming the harmony evaluation coefficient corresponding to the target music file, specifically includes:

[0071] In the preset time window, the pitch data corresponding to the target music file is collected in real time, and then the pitch change curve over time is confirmed. ;

[0072] By formula , and obtain the harmony evaluation coefficient corresponding to the target music file ,in, Represents a preset reference pitch change curve.

[0073] It should be noted that determining the comprehensive evaluation coefficient corresponding to the target music file specifically includes:

[0074] The spectral evaluation coefficient corresponding to the target music file , the rhythm evaluation coefficient corresponding to the target music file Harmony evaluation coefficient corresponding to the target music file Substitute into the formula , and obtain the comprehensive evaluation coefficient corresponding to the target music file ,in, They are respectively expressed as weight coefficients corresponding to the spectrum evaluation coefficient, rhythm evaluation coefficient, and harmony evaluation coefficient.

[0075] Further, searching a preset music database according to the comprehensive evaluation coefficient corresponding to the target music file, and then matching and confirming the corresponding target music file, specifically includes:

[0076] The comprehensive evaluation coefficient corresponding to the target music file Compare with the preset comprehensive evaluation interval to confirm the comprehensive evaluation coefficient corresponding to the target music file The target comprehensive assessment range;

[0077] Obtaining the comprehensive evaluation coefficient corresponding to each music file in the music database, screening out the music files whose comprehensive evaluation coefficients are within the target comprehensive evaluation interval, and storing the music files in the music file set;

[0078] Specifically, in the embodiment of the present application, the comprehensive evaluation coefficient corresponding to each music file in the music database is preferentially identified, and then the music files whose comprehensive evaluation coefficients corresponding to each music file in the music database are within the target comprehensive evaluation range are screened out.

[0079] Obtaining the number of elements in the music file set, and if the number of elements in the music file set is equal to a preset number threshold, determining that the music file in the music file set is the corresponding target music file;

[0080] If the number of elements in the music file set exceeds a preset number threshold, each element in the music file set is judged based on the first search method, and the corresponding target music file is matched and confirmed based on the judgment result.

[0081] Through the above technical scheme, by screening out music files whose comprehensive evaluation coefficients are within the target comprehensive evaluation range and storing the music files in a music file set, the number of elements in the music file set is further confirmed, and the corresponding target music files are judged based on the number of elements in the music file set, thereby effectively improving the accuracy of retrieval.

[0082] Specifically, the first search method is set to obtain the word count feature value and text feature value corresponding to each element in the music file set, and preferentially compare the word count feature value corresponding to each element in the music file set with the word count feature value corresponding to the target music file, and confirm the consistency. If they are inconsistent, further confirm the text feature value, compare the text feature value corresponding to each element in the music file set with the text feature value corresponding to the target music file, confirm the text feature ratio, and determine the element whose text feature ratio exceeds the preset threshold as the target music file.

[0083] Example 2

[0084] The embodiment of the present application also discloses a music database retrieval system based on feature extraction.

[0085] Reference Figure 2 , a music database retrieval system based on feature extraction, comprising:

[0086] A feature acquisition module, used to acquire feature data corresponding to a target music file, wherein the feature data includes spectrum data, rhythm data and harmony data;

[0087] A spectrum data analysis module, used to collect and analyze the spectrum data corresponding to the target music file, and then determine the spectrum evaluation coefficient corresponding to the target music file;

[0088] A rhythm data analysis module, used to collect and analyze the rhythm data corresponding to the target music file, and then determine the rhythm evaluation coefficient corresponding to the target music file;

[0089] A harmony data analysis module, used to collect and analyze the harmony data corresponding to the target music file, and then determine the harmony evaluation coefficient corresponding to the target music file;

[0090] The retrieval and matching module is used to comprehensively analyze the spectral evaluation coefficient, rhythm evaluation coefficient and harmony evaluation coefficient corresponding to the target music file, confirm the comprehensive evaluation coefficient corresponding to the target music file, and search the preset music database according to the comprehensive evaluation coefficient corresponding to the target music file, and then match and confirm the corresponding target music file.

[0091] The above contents are merely examples and explanations of the concept of the present invention. The technicians in this technical field may make various modifications or additions to the specific embodiments described or replace them in a similar manner. As long as they do not deviate from the concept of the invention or exceed the scope defined by the claims, they should all fall within the protection scope of the present invention.

[0092] In the description of this specification, the description with reference to the terms "one embodiment", "example", "specific example", etc. means that the specific features, structures, materials or characteristics described in conjunction with the embodiment or example are included in at least one embodiment or example of the present invention. In this specification, the schematic representation of the above terms does not necessarily refer to the same embodiment or example. Moreover, the specific features, structures, materials or characteristics described can be combined in any one or more embodiments or examples in a suitable manner.

[0093] The preferred embodiments of the present invention disclosed above are only used to help explain the present invention. The preferred embodiments do not describe all the details in detail, nor do they limit the invention to only specific implementation methods. Obviously, many modifications and changes can be made according to the content of this specification. This specification selects and specifically describes these embodiments in order to better explain the principles and practical applications of the present invention, so that those skilled in the art can understand and use the present invention well. The present invention is limited only by the claims and their full scope and equivalents.

Claims

1. A music database retrieval method based on feature extraction, It is characterized in that The following steps are involved: Acquire feature data corresponding to the target music file, wherein the feature data includes spectrum data, rhythm data and harmony data; Collecting and analyzing the spectrum data corresponding to the target music file, and then confirming the spectrum evaluation coefficient corresponding to the target music file; Collecting and analyzing the rhythm data corresponding to the target music file, and then confirming the rhythm evaluation coefficient corresponding to the target music file; Collecting and analyzing harmony data corresponding to the target music file, and then determining a harmony evaluation coefficient corresponding to the target music file; A comprehensive analysis is performed on the spectrum evaluation coefficient, rhythm evaluation coefficient and harmony evaluation coefficient corresponding to the target music file to confirm the comprehensive evaluation coefficient corresponding to the target music file, and a preset music database is searched based on the comprehensive evaluation coefficient corresponding to the target music file to match and confirm the corresponding target music file.

2. A music database retrieval method based on feature extraction according to claim 1, It is characterized in that The spectrum data includes spectrum, spectrum centroid data and spectrum bandwidth data, the rhythm data includes beat number, beat speed and beat intensity, and the harmony data includes pitch data.

3. A music database retrieval method based on feature extraction according to claim 2, It is characterized in that Collecting and analyzing the spectrum data corresponding to the target music file, and then confirming the spectrum evaluation coefficient corresponding to the target music file, specifically includes: Extracting frequency component data from the frequency spectrum corresponding to the frequency spectrum data, and normalizing the frequency component data, wherein the frequency component data includes the number of frequency components, the frequency and amplitude corresponding to each frequency component; By calculating the formula , get the spectrum centroid data corresponding to the target music file ,in, It is represented by the number corresponding to each frequency component. , Respectively expressed as The frequency and amplitude corresponding to each frequency component; By calculating the formula , get the spectrum bandwidth data corresponding to the target music file ; The spectral centroid data corresponding to the target music file Spectral bandwidth data corresponding to the target music file Substitute into the formula , and obtain the spectrum evaluation coefficient corresponding to the target music file ,in, Respectively represent the preset reference spectrum centroid data and reference spectrum bandwidth data, They are respectively expressed as the weight coefficients corresponding to the spectrum centroid data and the spectrum bandwidth data.

4. A music database retrieval method based on feature extraction according to claim 2, It is characterized in that Collecting and analyzing the rhythm data corresponding to the target music file, and then confirming the rhythm evaluation coefficient corresponding to the target music file, specifically includes: Extract the number of beats corresponding to the target music file from the rhythm data corresponding to the target music file , , and then through the formula Determine the duration of each beat ,in, Respectively expressed as The start time corresponding to each beat; By formula , get the tempo of the target music file , Expressed as The beat speed correction coefficient corresponding to each beat; Obtain the audio signal corresponding to each beat, and confirm the beat intensity corresponding to each beat based on the audio signal corresponding to each beat and the short-time energy method ; By calculating the formula , and obtain the rhythm evaluation coefficient corresponding to the target music file ,in, They represent the preset reference tempo speed and reference tempo intensity respectively. They are respectively expressed as the weight coefficients corresponding to the beat speed and beat intensity.

5. A music database retrieval method based on feature extraction according to claim 2, It is characterized in that Collecting and analyzing the harmony data corresponding to the target music file, and then determining the harmony evaluation coefficient corresponding to the target music file, specifically includes: In the preset time window, the pitch data corresponding to the target music file is collected in real time, and then the pitch change curve over time is confirmed. ; By formula , and obtain the harmony evaluation coefficient corresponding to the target music file ,in, Represents a preset reference pitch change curve.

6. A music database retrieval method based on feature extraction according to claim 5, It is characterized in that Confirming the comprehensive evaluation coefficient corresponding to the target music file includes: The spectral evaluation coefficient corresponding to the target music file , the rhythm evaluation coefficient corresponding to the target music file Harmony evaluation coefficient corresponding to the target music file Substitute into the formula , and obtain the comprehensive evaluation coefficient corresponding to the target music file ,in, They are respectively expressed as weight coefficients corresponding to the spectrum evaluation coefficient, rhythm evaluation coefficient, and harmony evaluation coefficient.

7. A music database retrieval method based on feature extraction according to claim 6, It is characterized in that The preset music database is searched according to the comprehensive evaluation coefficient corresponding to the target music file, and then the corresponding target music file is matched and confirmed, specifically including: The comprehensive evaluation coefficient corresponding to the target music file Compare with the preset comprehensive evaluation interval to confirm the comprehensive evaluation coefficient corresponding to the target music file The target comprehensive assessment range; Obtaining the comprehensive evaluation coefficient corresponding to each music file in the music database, screening out the music files whose comprehensive evaluation coefficients are within the target comprehensive evaluation interval, and storing the music files in the music file set; Obtaining the number of elements in the music file set, and if the number of elements in the music file set is equal to a preset number threshold, determining that the music file in the music file set is the corresponding target music file; If the number of elements in the music file set exceeds a preset number threshold, each element in the music file set is judged based on the first search method, and the corresponding target music file is matched and confirmed based on the judgment result.

8. A music database retrieval system based on feature extraction, It is characterized in that include: A feature acquisition module, used to acquire feature data corresponding to a target music file, wherein the feature data includes spectrum data, rhythm data and harmony data; A spectrum data analysis module, used to collect and analyze the spectrum data corresponding to the target music file, and then determine the spectrum evaluation coefficient corresponding to the target music file; A rhythm data analysis module, used to collect and analyze the rhythm data corresponding to the target music file, and then determine the rhythm evaluation coefficient corresponding to the target music file; A harmony data analysis module, used to collect and analyze the harmony data corresponding to the target music file, and then determine the harmony evaluation coefficient corresponding to the target music file; The retrieval and matching module is used to comprehensively analyze the spectral evaluation coefficient, rhythm evaluation coefficient and harmony evaluation coefficient corresponding to the target music file, confirm the comprehensive evaluation coefficient corresponding to the target music file, and search the preset music database according to the comprehensive evaluation coefficient corresponding to the target music file, and then match and confirm the corresponding target music file.

9. A computer-readable storage medium, Features: Instructions are stored, and when the instructions are executed on a computer, the computer is caused to execute a music database retrieval method based on feature extraction as described in any one of claims 1 to 7.

Images