JP5189952B2 - Music score recognition apparatus and computer program - Google Patents

Description

  The present invention relates to a score recognition apparatus and a computer program, and is particularly suitable for use in recognizing a score on paper.

Conventionally, a musical score on a paper surface is read using a scanner or the like, and a musical score data, musical notes, and other various musical symbols are recognized from the read musical score data, and performance data such as MIDI file data and display are displayed. Data for printing / printing is generated.
Specifically, when recognizing various musical symbols, the staff and paragraphs are recognized, and then the staff is erased, and then a reading label having an arbitrary rectangular shape is set for each symbol. The label data and the data registered in the dictionary in advance are subjected to pattern matching (see Patent Document 1). Here, in the case of a monochrome image (monochrome binary image), a label refers to a lump of black pixels that are four-connected or eight-connected, for example.

Japanese Patent Laid-Open No. 11-161738

  However, in the conventional technique described above, music symbols are recognized only from the result of pattern matching, so there is a possibility that the music symbols cannot be accurately recognized. For example, in tablature, note numbers are shown on horizontal lines corresponding to strings (in the following explanation, these lines are also called staves as necessary), but numbers are shown on the lines. There is a possibility that this number cannot be accurately recognized only from the result of pattern matching.

  The present invention has been made in view of such problems, and an object thereof is to make it possible to recognize a score including a tablature more accurately than in the past.

  The musical score recognition apparatus according to the present invention includes an acquisition unit that acquires, from an image reading unit, image information of a musical score including a tablature and a staff score that is paired with the tablature, a graphic written in the musical score, Recognizing means for recognizing to which musical symbol the graphic belongs by matching with a set dictionary, the note number of the note on the tablature, and the music notation paired with the note on the tablature Calculating means for calculating a difference between the note number and the note number calculated by the calculating means, and correcting means for correcting the music symbol recognized by the recognition means based on the difference between the note numbers calculated by the calculating means. And

  The computer program of the present invention includes an acquisition step of acquiring, from an image reading means, image information of a score including a tablature and a staff score paired with the tablature, a figure written in the score, and a preset A step of recognizing the musical symbol to which the figure belongs by matching with a dictionary, and the note number of the note on the tablature and the note on the staff notated in the tablature Causing the computer to execute a calculation step of calculating a difference between the note number of the note and a correction step of correcting the music symbol recognized by the recognition means based on the difference of the note number calculated in the calculation step. It is characterized by.

  According to the present invention, the music symbol recognized by matching is corrected based on the difference between the note number of the note on the tablature and the note number of the note paired with the note on the tablature. . Therefore, the score including the tablature can be recognized more accurately than before.

Hereinafter, an embodiment of the present invention will be described with reference to the drawings.
FIG. 1 is a block diagram illustrating an example of a configuration of a score recognition apparatus.
In FIG. 1, the score recognition apparatus 100 includes a CPU 1, ROM 2, RAM 3, HDD 4, FDD 5, MIDI interface circuit 6, scanner interface circuit 7, printer interface circuit 8, keyboard interface circuit 9, CRT interface circuit 10, and bus 15. doing. The scanner 11, printer 12, keyboard 13, and CRT 14 are connected to the score recognition apparatus 100.

As shown in FIG. 1, a score recognition apparatus 100 of this embodiment is obtained by adding a MIDI interface circuit 6 to a general computer system such as a personal computer.
The CPU 1 is a central processing unit that performs overall control of the score recognition apparatus 100 by developing a program stored in the ROM 2 on the RAM 3 and executing the program. The musical score recognition apparatus 100 has a built-in timer circuit that interrupts the CPU 1 at a predetermined cycle set in advance. The RAM 3 is used as an image data buffer, a work area, etc. in addition to a program area. An HDD (hard disk drive) 4 and an FDD (flexible disk drive) 5 store programs, image data, performance data, and the like.

  The CRT 14 displays video information output from the CRT interface circuit 10 based on the control of the CPU 1. Information input from the keyboard 13 is taken into the CPU 1 through the keyboard interface circuit 9. The printer 12 prints print information output from the printer interface circuit 8 based on the control of the CPU 1.

  The scanner 11 optically scans a score printed on paper and converts it into binary, gray scale, or color image data. As the scanner 11, an arbitrary type such as a flat bed type, a handy type, or a feeder type can be used. Image information read by the scanner 11 is taken into the RAM 3 or the HDD 4 via the scanner interface circuit 7. The MIDI interface circuit 6 is a circuit that transmits / receives MIDI data to / from an external MIDI device such as a sound module. The bus 15 connects each circuit in the score recognition apparatus 100. In addition, the score recognition apparatus 100 may include a pointing device such as a mouse, a serial interface circuit such as RS232C, and the like.

FIG. 2 is a main flowchart for explaining an example of processing of the CPU 1. It should be noted that here, description will be made mainly on matters necessary for describing the embodiment of the present invention, and detailed explanation of other matters will be omitted.
First, in step S1, the CPU 1 performs an image capturing process. In this image capturing process, a process of capturing image data of a score read by the scanner 12 into the RAM 3 is performed. FIG. 3 is a diagram illustrating an example of a score that is taken into the score recognition apparatus 100. Specifically, FIG. 3A is a diagram showing a part of a score in which a staff score and a tab score are paired. FIGS. 3B to 3D are diagrams conceptually showing first to third examples of recognition results by pattern matching in the region 303 of FIG.
As shown in FIG. 3A, in this embodiment, a musical score in which a staff 301 and a tablature 302 are paired is captured.

  Next, in step S2, the CPU 1 performs a staff recognition process. In the staff recognition processing, staff scanning start position detection processing, staff shift amount detection processing, and the like are performed. The outline of the staff scanning start position detection process is as follows. At a certain position in the horizontal (x) axis direction, the width of the black pixel and the width of the white pixel are obtained in order, and arranged in a staff based on the obtained width Detecting the position where it is coming in consideration of some error. Then, in order to remove the influence of the additional line (the horizontal line added to describe the note protruding from the staff), the condition that there is a white pixel width larger than the distance between the staffs on both sides of the staff array. Add. The midpoint of each black run in the horizontal (x) direction where black-and-white pixels are arranged that meet this condition is set as the staff scanning start position. Here, in the case of a monochrome image (monochrome binary image), the run means, for example, a length in which black pixels or white pixels are connected vertically or horizontally.

  Next, an outline of the staff shift amount detection process will be described. First, from the staff scanning start position (5 black pixel positions), the position is changed to the left and right one dot at a time, and when the number of black pixels becomes less than a certain number (for example, 3 or 4) out of 5 points In addition, the number of black pixels is checked by shifting the five points up and down, and the y coordinate is shifted in a direction in which the ratio of black pixels increases in the vertical (y) direction. The shift amount from the staff scanning start position is defined as the staff shift amount. Then, the staff is detected by scanning to the position where the number of black pixels becomes zero in the left-right direction (horizontal direction) of the staff scanning start position.

  Next, in step S3, the CPU 1 performs a paragraph recognition process. In the paragraph recognition process, a paragraph recognition process and a bracket recognition process are performed. In the paragraph recognition process, the staff is detected in the entire image, a set of staffs in which the left ends of the staffs are almost in the same place is searched, and it is inspected whether the ends of the staff are connected with black pixels. And recognize paragraphs. When staffs are connected with brackets, there may be notes that cross between staffs, so it is better to process staffs connected with brackets in one unit. . The brackets are recognized within a predetermined range to the left of the paragraph line. Here, the paragraph line refers to a line that connects the left ends of the staffs of a plurality of parts that are proceeding simultaneously, and the paragraph refers to a lump of a plurality of staffs that are connected by paragraph lines. When there is one part, a block corresponding to a plurality of part paragraphs is called a paragraph. Also, the brackets are music symbols that are often used for piano scores, in particular, parentheses indicated by “[” for collecting two or three staffs.

  Next, in step S4, the CPU 1 performs part recognition processing. In the part recognition process, each part is recognized based on the results recognized in the staff recognition process and the paragraph recognition process. In the part recognition processing, processing rectangle determination processing, staff erasing processing, music symbol recognition processing, and the like are performed. In the process rectangle determination process, a rectangle that is fairly wide including the found staff is taken and is used as a recognition process rectangle. The size of the rectangle is not less than the maximum area where the music symbols related to the staff are present, and is adjusted so that the necessary music symbols are not erased by correcting the staff inclination. Subsequent recognition is performed within this rectangle.

In the staff erasing process, a portion of the obtained staff whose line width is equal to or smaller than a threshold (a portion not overlapping with a music symbol) is erased.
In the music symbol recognition process, a rectangular label is set so as to touch the outer edge of continuous black pixels for the musical score image data from which the staff has been deleted, and the black pixel data in the rectangular label is registered in the dictionary in advance. Various music symbols are recognized by performing pattern matching with the recorded data. By this music symbol recognition processing, the numbers written on the tablature are recognized. In pattern matching, in addition to template matching, the size of a rectangular label and a weight direction index can be used, or a peripheral feature can be used.

Next, in step S5, the CPU 1 performs a combination process. In the merging process, the score recognized by the staff recognition process, the paragraph recognized by the paragraph recognition process, the contents of the part recognized by the part recognition process, etc. are combined to generate music score data. Do.
Next, in step S6, the CPU 1 performs recognition result correction processing. In the recognition result correction process, the result of the pattern recognition in the music symbol recognition process is used as the note number of the notation of the staff 301 and the note of the tab staff 302 (the position of the staff on which the fret number and the fret number are written). Correct the note based on the difference from the note number. Details of the recognition result correction processing will be described later.

In the present embodiment, a part template including setting information that needs to be set for the score after the recognition of the score is stored in the HDD 4 or the like in advance. Then, before the recognition result correction process in step S6 described above is performed, a part template corresponding to each part of the score captured in step S1 is extracted from the plurality of part templates. More specifically, a character string dictionary that recognizes characters constituting a part name (part name) written on the left side of a paragraph, etc. by performing symbol matching and assumes the sequence of the characters as a part name. The character string indicating the part name is recognized from the matching result. Then, a part template representing an instrument corresponding to the recognized character string is selected from a plurality of part templates (acquisition instrument ID corresponding to the recognized character string is obtained, and the part template corresponding to the instrument ID is selected as Select from multiple part templates).
FIG. 4 is a diagram illustrating an example of a part template. In FIG. 4, a part template 401 includes a template name 402, a part name 403, a part name shortening 404, a transposing instrument setting 405, a clef 406, a tab score 407, and tab score tuning information 408.

The template name 402 stores, for example, a Japanese character string or an alphabetic character string as a character string for identifying the part template 401.
The part name 403 stores a character string for identifying a part to which the part template 401 is applied.
The part name shortening 404 stores a character string for identifying the short notation of the part to which the part template 401 is applied.
The transposing instrument setting 405 is used to identify whether or not the instrument to which the part template 401 is applied is a transposing instrument, and if the instrument is a transposing instrument, what the transposing instrument is. Numeric values are stored. In this embodiment, when the instrument to which the part template 401 is applied is not a transposing instrument, the transposing instrument setting 405 stores 0. In this embodiment, when the musical instrument to which the part template 401 is applied is a trumpet, a wood bass, a piccolo, or an alto saxophone, -2, -12, 12, and -9 are stored in the transposing instrument setting 405, respectively. Is done.

The clef 406 stores a numerical value corresponding to the type of clef to which the part template 401 is applied. In the present embodiment, when the clef to which the part template 401 is applied is a treble clef, 0 is stored in the clef 406. In this embodiment, when the clef to which the part template 401 is applied is a clef, 1 is stored in the clef 406.
The number of tablature lines 407 stores a numerical value representing the “number of musical score lines (number of staffs)” to which the part template 401 is applied.
The tablature tuning information 408 stores numerical values representing tuning for each line of the score. The tablature tuning information 408 stores information for “number of lines of musical score” to which the part template 401 is applied. FIG. 5 is a diagram illustrating an example of the tablature tuning information 408.

  The part template 401 only needs to include at least transposition instrument setting 405, clef 406, tab score 407, and tab score tuning information 408. However, it is preferable that the template name 402 and the part name 403 are included. Also, the information stored in the template name 402, part name 403, part name shortening 404, transposing instrument setting 405, clef 406, tab score 407, and tab score tuning information 408 are limited to those described above. It is not something.

  Then, the CPU 1 selects one part template 401 corresponding to the number of lines recognized in the part recognition process (the number of horizontal lines corresponding to the strings of the tablature) from the selected part templates 401, and the image of step S1. The part template 401 is applied to the tablature 302 of the musical score captured by the data capturing process.

  The contents of the part template 401 as described above and the part template 401 applied to the tablature 302 are based on, for example, the user's operation on the “paragraph correction screen” for changing the setting of the staff or paragraph. Can be changed.

Returning to the description of FIG. 2, in step S <b> 7, the CPU 1 performs other processing. In other processing, processing such as generating performance data such as MIDI file data is performed.
As described above, actually, various processes are executed in addition to the above process. However, since the process other than the recognition result correction process in step S6 can be realized by a known technique, the details are described here. The detailed explanation is omitted.

Next, an example of the recognition result correction process in step S6 of FIG. 2 will be described with reference to the flowchart of FIG.
First, in step S11, the CPU 1 calculates the note number of the note of the staff 301 based on the content of the music symbol recognized by the part recognition process, and stores the calculated note number as the note information. The calculation of the note number can be performed in the same way as when the note number is detected when playing the staff. However, in the present embodiment, the note number of each note is determined from the beginning of the score (song), taking into account only musical symbols related to the pitch (clef, key signature, accidental symbol, tie, ottava, etc.). calculate. Further, in the present embodiment, even if there is a repetition symbol such as “DS” or “coder”, the note number is calculated without following the repetition symbol. That is, the note number is detected without going backward from the beginning to the end of the score (song).

  In step S12, the CPU 1 calculates the note number of the note of the tablature 302 based on the content of the music symbol recognized by the part recognition process, and stores the calculated note number as the note information. The calculation of the note number can be performed in the same way as when the note number is detected when playing the tablature. However, in the present embodiment, the “note number (fret number)” written in the “position corresponding to the note head” of the tablature 302 is included in the part template 401 applied to the tablature 302. The tab score tuning information 408, which is obtained by adding the information corresponding to the string on which the fret number is written, is used as the note number. Here, note heads are those corresponding to “black balls with notes shorter than quarter notes or perforated balls with notes longer than half notes” on the staff.

  Next, in step S13, the CPU 1 performs a previous part reference information acquisition process. In the previous part reference information acquisition process, information on the staff 301 that is paired with the tablature 302 is referred to, and based on the result of the reference, the musical symbol that is paired with the note is added to the information of each note on the tablature 302. This is a process for storing the index information. The index information includes an index number assigned to each musical symbol from the top of the staff 301 and the information (note number etc.) indicating the contents of the musical symbol. Details of this previous part reference information acquisition processing will be described later.

  Next, in step S14, the CPU 1 determines whether or not the number (fret number) of the tablature 302 is to be corrected. For example, when the staff 301 that is paired with the tablature 302 does not exist, it can be determined that the number (fret number) of the tablature 302 is not corrected. Further, when the number or ratio of the notes on the tablature 302 that can detect the paired music symbol is equal to or greater than the threshold value, it is determined that the number (fret number) of the tablature 302 is to be corrected. It can be determined that the numeral 302 (fret number) is not corrected. In addition, when the number or ratio of “musical symbols on the staff 301” that are not paired with the notes on the tablature 302 side is equal to or less than the threshold, it is determined that the number (fret number) of the tablature 302 is corrected. Otherwise, it can be determined that the number (fret number) of the tablature 302 is not corrected. In the above, the ratio is, for example, a value obtained by dividing the number of corresponding notes by the total number of notes in the staff 301 or the tablature 302.

As a result of the determination, if the number (fret number) of the tablature 302 is not corrected, the process proceeds to step S24, and after other processes are performed, the process according to the flowchart of FIG. On the other hand, when correcting the number (fret number) of the tablature 302, the process proceeds to step S15.
In step S15, the CPU 1 creates a note number comparison table.
FIG. 7 is a diagram illustrating an example of a note number comparison table. Specifically, FIG. 7A is a diagram showing an example of the note number comparison table created in step S15 of FIG. 6, and FIG. 7B is a note created in step S19 of FIG. It is a figure which shows an example of a number comparison table.

  First, based on the note number of the note of the tablature 302 and the “index information of the note of the staff 301” stored in the note information, the CPU 1 calculates the note and the note number from the note number of the note of the tablature 302. A value obtained by subtracting the note numbers of the “staff 301 notes” to be paired is calculated for all the notes on the tablature 302. If the note information of the tablature 302 is not stored in the note information of the tablature 302, the note number difference is set to the normal number so that this state can be determined for the notes of the tablature 302. It is assumed that the value is not possible (for example, “−1000”). Further, in the following description, a value obtained by subtracting the note number of the note of the tablature 302 from the note number of the “stave 301 note” paired with the note is referred to as a note number difference as necessary.

After that, the CPU 1 calculates the maximum value of the numbers having the same note number difference for each measure, and stores the calculated value in the area 701a of the note number comparison table 701 shown in FIG. For example, the number of notes in string 1 of measure 1 of tablature 302 is 4, and there are notes in paired music notation 301 in the notation 301, and the difference between the note numbers of these four notes is “1”. , “1”, “2”, “0”. In this case, “1” is stored as a representative value in the “cell specified by the bar 1 and the string 1” in the area 701a.
When the values are stored in all the cells in the area 701a as described above, the value of each cell is extracted for each string with a value of 0 or more, and the maximum number of the extracted values having the same value is extracted. The value is calculated, and the calculated value is stored in an area 701b of the note number comparison table 701 shown in FIG. For example, suppose N is 4, and the values stored in measures 1 to 4 in string 1 are “1”, “1”, “0”, and “1”, respectively. In this case, “1” is stored as a representative value in the cell corresponding to the string 1 in the region 701b.

When there are a plurality of maximum values of the same number of note numbers, the smaller value is selected, for example.
The number of note number differences calculated to determine the value to be stored in the cell in the area 701a (that is, the number of “notes in the staff 301 and tab 302” in the bar corresponding to the cell in the area 701a) is calculated. When the value is equal to or less than the threshold value, the value of the cell may be blank. In this way, it is possible to reduce the influence of pattern matching erroneous recognition in the staff 301. In this case, the number of note number differences may be stored in each cell so that the number of note number differences in each cell can be easily determined.

Next, in step S16, the CPU 1 performs tuning information correction processing. In this tuning information correction process, it is determined whether or not to correct the tab score tuning information 408 in the part template 401 applied to the tab score 302 based on the note number comparison table 701 created in step S15. If it is necessary to correct it, correct it. Details of this tuning information correction processing will be described later.
Next, in step S17, the CPU 1 determines whether or not the tablature tuning information 408 has been changed as a result of the tuning information correction processing in step S16. If the result of this determination is that the tablature tuning information 408 has been changed, the process proceeds to step S18. In step S18, the CPU 1 recalculates the note numbers of the notes of the tablature 302 based on the changed tablature tuning information 408. Then, the process proceeds to step S19.
On the other hand, if the tablature tuning information 408 has not been changed, step S18 is omitted and the process proceeds to step S19.

In step S19, the CPU 1 creates a note number comparison table as shown in FIG.
First, as in the case of creating the note number comparison table 701 shown in FIG. 7A, the CPU 1 determines the maximum number of cells having the same note number difference in the cells specified by each string and each measure. Store the value.
Thereafter, the CPU 1 calculates the average value of the values stored in all the string cells for each bar, and stores the calculated value in the area 702a of the note number comparison table 702. For example, when the values of the strings 1 to 6 in the bar 1 are “1”, “1”, “2”, “2”, “3”, and “3”, they correspond to the bar 1 in the area 702a. “2 (= ((1 + 1 + 2 + 2 + 3 + 3) / 6))” is stored in the cell.

  Then, the CPU 1 determines whether or not there are consecutive bars whose values stored in the region 702a are equal to or greater than the threshold value as described above. As a result of this determination, if there are consecutive bars whose values stored in the region 702a are equal to or greater than the threshold, the average value of the values stored for the cells of the continuous bars is calculated for each string. To do. When the measure whose value stored in the region 702a is equal to or greater than the threshold is the measures 10-15, the average value for each string of the values of the measures 10-15 displayed in gray is stored in the region 702b.

  Next, in step S20, the CPU 1 determines whether or not a clef erroneous recognition area has been detected based on the note number comparison table 702 created in step S19. Here, the clef symbol misrecognition region refers to a “stave 301 region” in which a clef symbol is recognized but no clef symbol is present in the pattern matching. Specifically, for example, a predetermined region at the beginning of the first measure in the measure for which the region 702b is calculated (the measure corresponding to the portion displayed in gray in FIG. 7B) is a clef symbol misrecognition region. It becomes.

  A specific example of the processing in step S20 will be described. First, the CPU 1 refers to the note number comparison table 702 created in step S19, and varies the values stored in the area 702b (difference between the maximum value and the minimum value). Or standard deviation) is determined to be less than or equal to a threshold value (within a predetermined range). As a result of the determination, when the variation in the value stored in the area 702b is equal to or less than the threshold value, it is determined that the clef misrecognition area is detected, and if not, the clef misrecognition area is not detected. judge.

If the result of this determination is that a clef misrecognition area has not been detected, steps S21 and S22 are omitted and the process proceeds to step S23, which will be described later. On the other hand, if a clef erroneous recognition area is detected, the process proceeds to step S21.
In step S21, the CPU 1 changes the result of recognition by pattern matching of the staff 301. A specific example of the process in step S21 will be described. First, the CPU 1 calculates the average pitch of the bars in which the area 702b is calculated (the bar corresponding to the portion displayed in gray in FIG. 7B). Then, the CPU 1 applies to the measure (the measure corresponding to the portion displayed in gray in FIG. 7B) based on the average and the clef applied to the measure before the measure. Estimate the clef to be played. The CPU 1 adds the estimated clef to the clef recognition area.

  Note that, as a result of pattern matching performed in the music symbol recognition process in the part recognition process of step S4 in FIG. 2, the music symbol having the highest matching degree (score) among the music symbols whose matching degree (score) is not equal to or greater than the threshold value. When the next point ID for identification is stored in the clef recognition area, the clef estimated as described above matches the music symbol identified by the next point ID. Only the estimated clef may be added to the clef recognition area. Here, the next point ID is used to identify a music symbol that is most likely to be applicable among music symbols that are not recognized because the degree of matching (score) is not equal to or greater than a threshold in pattern matching.

Next, in step S22, CPU1 changes the information regarding the pitch of the note of the bar to which the clef is applied based on the clef changed in step S21. Then, the process proceeds to step S23.
In step S23, the CPU 1 performs a tab score correction process. In this tablature number correction process, if the difference between the note number of the note on the staff 301 and the note number of the note on the tablature 302 is greater than or equal to the threshold value, the note number of the tab staff is added to the note number of the note on the staff 301. Change the number together. Details of the tablature number correction process will be described later.
Next, in step S24, the CPU 1 performs other processing and ends the processing according to the flowchart of FIG.

Next, an example of the previous part reference information acquisition process in step S13 in FIG. 6 will be described with reference to the flowchart in FIG.
First, in step S101, the CPU 1 substitutes the variable nMeasureT for the measure number to which the note to be processed of the tablature 302 belongs.
Next, in step S102, the CPU 1 in the horizontal direction from the right end of the measure of the tablature 302 specified by the variable nMeasureT (the position of the bar line after the measure) to the processing target note of the note tablature 302. The shortest distance is obtained, and the obtained distance is substituted into the variable xT. In the description of the flowchart of FIG. 8, the shortest distance in the horizontal direction is referred to as the shortest distance as necessary.

Next, in step S103, the CPU 1 starts from the right end of the bar of the staff 301 that is the bar of the staff 301 specified by the variable nMeasureT and that is paired with the tablature 302 to which the note to be processed belongs. A note whose shortest distance xB is closest to the variable xT is detected.
Next, in step S <b> 104, the CPU 1 determines whether or not a note whose absolute value (= | xT−xB |) of the difference between the variable xT and the shortest distance xB is equal to or less than a threshold is detected from the staff 301. As a result of the determination, if a note whose absolute value (= | xT−xB |) of the difference between the variable xT and the shortest distance xB is not greater than the threshold is not detected from the staff 301, the process proceeds to step S110 described later. On the other hand, if a note whose absolute value (= | xT−xB |) of the difference between the variable xT and the shortest distance xB is less than or equal to the threshold is detected from the staff 301, the process proceeds to step S105.

In step S105, the CPU 1 determines whether or not at least one of the note to be processed in the tablature 302 and the note paired with the note in the staff 301 is a chord. As a result of the determination, if both the note to be processed in the tablature 302 and the note paired with the note in the staff 301 are not chords, the process proceeds to step S111 described later.
On the other hand, if at least one of the note to be processed in the tablature 302 and the note paired with the note in the staff 301 is a chord, the process proceeds to step S106.

In step S106, the CPU 1 selects a combination of notes that minimizes the difference between the note number of the note to be processed in the tablature 302 and the note number of the note paired with the note in the staff 301, for example, a permutation.・ Detect by performing combination calculation.
For example, as shown in FIG. 3B, when the number of constituents of the chord to be processed in the tablature 302 is the same as the number of constituents of the chord paired with the chord in the staff 301, The notes 312 and 313 and the notes 310 and 311 of the staff 301 are combined. For example, the difference between the note number of the note 313 on the tablature 302 and the note number of the note 311 on the staff 301 is greater than the difference between the note number on the note 312 on the tablature 302 and the note number of the note 311 on the staff 301. The difference between the note number of the note 312 of the tablature 302 and the note number of the note 310 of the staff 301 is smaller than the difference between the note number of the note 313 of the tablature 302 and the note number of the note 310 of the staff 301. If it is smaller, the combination of the note 313 of the tablature 302 and the note 311 of the staff 301, the combination of the note 312 of the tablature 302 and the note 310 of the staff 301 are detected.

  Further, as shown in FIG. 3C, when the number of constituent sounds of the chord to be processed in the tablature 302 is smaller than the number of constituent sounds of the chord paired with the chord in the staff 301, the tablature 302 is displayed. Are combined with any one of the notes 314 and 315 of the staff 301. For example, the difference between the note number of the note 316 of the tablature 302 and the note number of the note 314 of the staff 301 is greater than the difference between the note number of the note 316 of the tablature 302 and the note number of the note 315 of the staff 301. If it is smaller, the combination of the note 316 on the tablature 302 and the note 314 on the staff 301 is detected.

  Further, as shown in FIG. 3D, when the number of constituent sounds of the chord to be processed in the tablature 302 is larger than the number of constituent sounds of the chord paired with the chord in the staff 301, the tablature 302 is displayed. Are combined with any one of the notes 318 and 319. For example, the difference between the note number of the note 319 of the tablature 302 and the note number of the note 317 of the staff 301 is greater than the difference between the note number of the note 318 of the tablature 302 and the note number of the note 317 of the staff 301. If it is smaller, the combination of the note 319 of the tablature 302 and the note 317 of the staff 301 is detected.

  As described above, when a note combination in which the difference between the note number of the note to be processed in the tablature 302 and the note number of the note paired with the note in the staff 301 is the smallest is detected, step S107. Proceed to In step S107, the CPU 1 adds the note index information of the staff 301 to the note information combined with the note of the tab staff 302 based on the detected combination.

Note that the note index information of the staff 301 is not necessarily added to the note information of the tab staff 302 as long as it is information that can identify the notes of the tab staff 302.
In this case, the index information of the note combined with the one constituent sound of the staff 301 can be added to the information of the constituent sound of the chord to be processed in the tablature 302. In this way, adding the index information of the notes combined with the one constituent sound of the staff 301 to the information of the constituent sound of the chord to be processed of the tablature 302 is the chord to be processed of the tablature 302. This can be realized regardless of whether the number of constituent sounds of the chord and the number of constituent sounds of the chord paired with the chord of the staff 301 are the same or different.

Next, in step S <b> 108, the CPU 1 assigns a processed flag to the information on the notes to be processed in the tablature 302.
Next, in step S109, the CPU 1 determines whether or not processing has been performed for all the notes of the tablature 302. If the result of this determination is that processing has not been performed for all the notes on the tablature 302, the next processing target note on the tablature 302 is selected and the processing from step S101 onward is performed. On the other hand, when all the notes on the tablature 302 have been processed, the processing according to the flowchart of FIG.

As described above, when it is determined in step S104 that a note whose absolute value (= | xT−xB |) of the difference between the variable xT and the shortest distance xB is equal to or less than the threshold is not detected from the staff 301, Proceed to step S110. When proceeding to step S110, the CPU 1 starts from the right end of the measure in the measure of the staff 301 that is paired with the tablature 302 to which the note to be processed belongs, which is the measure of the staff 301 specified by the variable nMeasureT. The “musical symbols other than notes” whose shortest distance xB is closest to the variable xT are detected. Then, the process proceeds to step S111.
In step S111, the CPU 1 uses the index information of the musical symbol other than the note or the note detected in step S103 or S110 as the information on "the musical note paired with the musical symbol other than the note or the musical note" in the tablature 302. Assign to. Then, the process proceeds to step S108 described above.

Next, an example of the tuning information correction process in step S16 in FIG. 6 will be described with reference to the flowchart in FIG.
First, in step S 201, it is determined whether or not the difference between the note number of the note on the tablature 302 and the note number of the note on the staff 301 that is paired with the tablature 302 can be regarded as “0”. A specific example of the process in step S201 will be described. When all the values stored in the area 701b of the note number comparison table 701 shown in FIG. It is determined that the difference between the number and the note number of the note of the staff 301 that is paired with the tablature 302 can be regarded as “0”, otherwise, it is determined that it cannot be regarded as “0”.

As a result of the determination, if the difference between the note number of the note on the tablature 302 and the note number of the note on the staff 301 that is paired with the tablature 302 can be regarded as “0”, the processing according to the flowchart of FIG. finish.
On the other hand, if the difference between the note number of the note on the tablature 302 and the note number of the note on the staff 301 that is paired with the tablature 302 cannot be regarded as “0”, the process proceeds to step S202. In step S202, the CPU 1 substitutes 1 for a variable i for identifying a string of the tablature 302.

  Next, in step S203, the CPU 1 determines whether or not it can be considered that the difference of the note numbers in the strings identified by the variable i is smaller than the threshold value Thr. A specific example of the process of step S203 will be described. When the value stored in the cell corresponding to the string identified by the variable i in the area 701b of the note number comparison table 701 is smaller than the threshold Thr, the variable It is determined that the note number difference in the string identified by i can be considered to be smaller than the threshold value Thr, and otherwise, it is determined that it cannot be considered. As the threshold Thr, 1 or 2 can be adopted. In this way, even if there is an error in the note number due to erroneous recognition of a temporary symbol or key signature by pattern matching, it is possible to appropriately perform processing.

As a result of this determination, if it cannot be considered that the difference of the note number in the string identified by the variable i is smaller than the threshold value Thr, the process skips step S204 and proceeds to step S205 described later. On the other hand, if the note number difference in the string identified by the variable i can be considered to be smaller than the threshold value Thr, the process proceeds to step S204.
In step S204, the CPU 1 sets the string information identified by the variable i to the string information identified by the variable i in the tab score tuning information 408 in the part template 401 applied to the tab score 302. Note number difference (value stored in the cell corresponding to the string identified by the variable i in the area 701b of the note number comparison table 701) is added.

  Here, in the area 701b of the note number comparison table 701, the value stored in the cell corresponding to the string identified by the variable i is added, but this is not necessarily required. For example, a note number difference and a value to be changed in the tablature tuning information 408 are stored in association with each other as a library. Then, the note number difference closest to the value stored in the “cell in region 702b” corresponding to the string identified by the variable i is searched from this library. Then, a value registered in the library in association with the searched note number difference may be set in the tablature tuning information 408.

  Next, in step S <b> 205, the CPU 1 determines whether or not the variable i is the total number of horizontal lines corresponding to the strings of the tablature 302. As a result of the determination, if the variable i is not the total number of horizontal lines corresponding to the strings of the tablature 302, the process proceeds to step S206, 1 is added to the variable i, and the process returns to step S203. On the other hand, if the variable i is the total number of horizontal lines corresponding to the strings of the tablature 302, the process proceeds to step S207. In the example shown in FIGS. 3 and 5, the total number of horizontal lines corresponding to the strings of the tablature 302 is “6”.

  In step S207, the CPU 1 determines whether or not the tablature tuning information 408 that has been corrected in step S204 has simply changed. Here, the simple change means that the value of the tablature tuning information 408 simply changes (decreases or increases) as the number of strings increases as shown in FIG. Even when the values of the plurality of strings in the tablature tuning information 408 are the same, it is assumed that the tablature tuning information 408 after the correction in step S204 does not change simply. As a result of this determination, if the tablature tuning information 408 that has been corrected in step S204 has simply changed, the process according to the flowchart of FIG. 9 ends.

  On the other hand, if the tablature tuning information 408 that has been corrected in step S204 has not simply changed, the process proceeds to step S208. In step S208, the CPU 1 determines whether to return the tablature tuning information 408 to the one before correction. A specific example of the process of step S208 will be described. When the number of strings in the tablature 302 is a number corresponding to an instrument (four-string ukulele) for which the tablature tuning information 408 is considered not to change simply, It is determined that the tablature tuning information 408 is not returned to the one before correction. Further, when the tablature 302 is not the number corresponding to an instrument for which the tablature tuning information 408 is considered not to change simply, the tablature tuning information 408 is changed to the one before correction. Determine to return.

  As a result of this determination, if the tablature tuning information 408 is not returned to the one before correction, the processing according to the flowchart of FIG. 9 is terminated. On the other hand, when returning the tablature tuning information 408 to the one before correction, the process proceeds to step S209. In step S209, the CPU 1 returns the tablature tuning information 408 corrected in step S204 to the value before correction. And the process by the flowchart of FIG. 9 is complete | finished.

Next, an example of the tab score correction process in step S23 of FIG. 6 will be described with reference to the flowchart of FIG.
First, the CPU 1 adds the note number NoteTab of the note in the tablature 302 and the note number NoteBef of the note in the staff 301 added to the note information in the previous part reference information acquisition process in step S13 (FIG. 8) in FIG. It is determined whether or not the absolute value of the difference from “is greater than the threshold NoteDiffThr. If the absolute value of the difference between the note number NoteTab of the note on the tablature 302 and the note number NoteBef of the note on the staff 301 is not greater than the threshold NoteDiffThr, the number of the tablature 302 is corrected. The process proceeds to step S311 to be described later. On the other hand, if the absolute value of the difference between the note number NoteTab of the note on the tablature 302 and the note number NoteBef of the note on the staff 301 is greater than the threshold NoteDiffThr, the process proceeds to step S302.

  Note that if the note of the tablature 302 is a chord and the information on the note contains only one piece of information on the chord that is paired with the chord, the part recognition process in step S4 in FIG. Based on the information indicating the appearance timing of notes obtained by the above method, the note number of the chord component sound of the staff 301 that is paired with the chord component sound of the tablature 302 is extracted. Then, as shown in FIGS. 3C and 3D, as a result of recognition by pattern matching in the part recognition process, the chord component of the tablature 302 and the chord of the stave 301 paired with the chord are obtained. When the constituent sounds are different, as described in step S106 in FIG. 8, for example, permutation / combination calculation is performed, and the note number of the note to be processed in the tablature 302 and the note in the staff 301 are A combination of notes with the smallest difference from the note number of the paired notes is detected. Then, a note number difference is calculated between the detected notes. Here, when there is no note of the staff 301 that is paired with the note of the tablature 302, the number of the note of the tablature 302 is not changed.

In step S302, the CPU 1 subtracts the tablature tuning information 408 (Tuning [Loc]) applied to the note to be processed in the tablature 302 from the note number NoteBef of the note in the staff 301. A new number NewNumTab of the note to be processed 302 is set. If the tab score tuning information 408 is corrected in step S16 (FIG. 9) in FIG. 6, the corrected tab score tuning information 408 is used in step S302.
Next, in step S303, the CPU 1 determines whether or not the new number NewNumTab of the note to be processed in the tablature 302 is greater than 0 and less than 28. If the result of this determination is that the new note number NewNoteTab of the note in the tablature 302 is greater than 0 and less than 28, step S304 is omitted and the process proceeds to step S305 described later.

On the other hand, if the new number NewNumTab of the note to be processed in the tablature 302 is not greater than 0 and not less than 28, the process proceeds to step S304. In step S304, the CPU 1 deletes the tens place value of the new number NewNumTab of the note to be processed of the tablature 302. Then, the process proceeds to step S305.
In step S305, the CPU 1 decomposes both the original number OldNumTab of the note to be processed in the tablature 302 and the new number NewNumTab of the note to be processed in the tablature 302 for each digit. For example, when the note number is 12, the note number is decomposed into 1 and 2.

Next, in step S306, the CPU 1 determines whether or not both the original number OldNumTab of the note to be processed of the tablature 302 and the new number NewNumTab of the note to be processed of the tablature 302 are two digits. . As a result of the determination, if both the original number OldNumTab of the note to be processed in the tablature 302 and the new number NewNumTab of the note to be processed in the tablature 302 are not two digits, the process proceeds to step S312 described later. .
On the other hand, if both the original number OldNumTab of the note to be processed in the tablature 302 and the new number NewNumTab of the note to be processed in the tablature 302 are two digits, the process proceeds to step S307. In step S307, the CPU 1 checks for each digit whether or not there is a corrected number corresponding to the number before correction in a table stored in advance in the HDD 4 or the like.

That is, in this table, the numbers before correction and the numbers after correction are stored in association with each other. Specifically, this table shows a number before correction and a number that may be erroneously recognized by symbol matching for the number before correction. Are stored in association with each other.
The CPU 1 determines that the corrected number associated with the uncorrected number that matches the value of the first digit OldNumTab of the note to be processed in the tablature 302 is the value of the note to be processed in the tablature 302. Determine if the new digit NewNumTab matches the one's place value. As a result of this determination, if they match, it is determined that there is a corrected number corresponding to the number before correction for the first place, and if these do not match, It is determined that there is no corrected number corresponding to the number before correction. The tenth place can be checked in the same way as the first place.
For example, the value of the original number OldNumTab of the note to be processed of the tablature 302 is 12, the value of the new number NewNumTab of the note to be processed of the tablature 302 is 21, and the table shows the number before correction. 1 and 2 are stored as corrected numbers. In this case, for the tens place, it is determined that there is a corrected number corresponding to the number before the correction, and for the first place, it is determined that there is no corrected number corresponding to the number before the correction. It is said.

Next, in step S308, the CPU 1 determines whether or not there is a corrected number corresponding to the number before correction in a table stored in advance in the HDD 4 or the like for all digits. As a result of this determination, if there is no corrected number corresponding to the number before correction in the table stored in advance in the HDD 4 or the like for all the digits, the process proceeds to step S311 described later.
On the other hand, if there is a corrected number corresponding to the number before correction in the table stored in advance in the HDD 4 or the like for all the digits, the process proceeds to step S309. In step S309, the absolute value of the difference between the new number NewNumTab of the note to be processed in the tablature 302 and the note number NoteBef "of the note paired with the note in the staff 301 is smaller than the threshold value TabDiffThr. It should be noted that “0”, “1”, “2” or the like can be adopted as the threshold TabDiffThr. In this way, when a value of “1” or more is adopted as the threshold value TabDiffThr, the number of the tablature 302 is unintentionally changed in response to a slight difference in the note number due to misidentification of a temporary symbol or the like. Can be prevented.

As a result of this determination, when the absolute value of the difference between the new number NewNumTab of the note to be processed of the tablature 302 and the note number NoteBef "of the note paired with the note of the staff 301 is smaller than the threshold value TabDiffThr Skips step S310 and proceeds to step S311 described below.
On the other hand, when the absolute value of the difference between the new number NewNumTab of the note to be processed in the tablature 302 and the note number NoteBef "of the note paired with the note in the staff 301 is not smaller than the threshold TabDiffThr, Proceeding to step S310, when proceeding to step S310, the CPU 1 adds a new number NewNumTab of the note to be processed of the tablature 302 to the "numbered tablature information" included in the information of the note to be processed of the tablature 302. Then, the process proceeds to step S311.

  In step S311, the CPU 1 determines whether or not the numbers of all the notes on the tablature 302 have been checked. If the result of this determination is that the numbers of all the notes on the tablature 302 have not been checked, the next note is designated and the processing from step S301 is performed again. On the other hand, when the numbers of all the notes on the tablature 302 are checked, the processing according to the flowchart of FIG.

  As described above, when it is determined in step S306 that both the original number OldNumTab of the note to be processed of the tablature 302 and the new number NewNumTab of the note to be processed of the tablature 302 are not two digits. Advances to step S312. In step S312, the CPU 1 determines whether both the original number OldNumTab of the note to be processed in the tablature 302 and the new number NewNumTab of the note to be processed in the tablature 302 are one digit. To do. As a result of the determination, if both the original number OldNumTab of the note to be processed of the tablature 302 and the new number NewNumTab of the note to be processed of the tablature 302 are one digit, the process proceeds to step S307 described above. move on. In this case, in step S307, only the first place is checked.

On the other hand, when both the original number OldNumTab of the note to be processed in the tablature 302 and the new number NewNumTab of the note to be processed in the tablature 302 are not one digit, one of them is two digits and the other Is one digit, and the process proceeds to step S313.
In step S313, the CPU 1 determines whether or not there is a corrected number corresponding to the number before correction in a table stored in advance in the HDD 4 or the like, as a one-digit value and a two-digit value. Check in between. That is, the CPU 1 determines that the corrected number associated with the number before correction that matches the value of the original digit OldNumTab of the note to be processed in the tablature 302 is the processing target of the tablature 302. Judges whether the new digit NewNumTab of the note matches the one's place value. As a result of this determination, if they match, it is determined that there is a corrected number corresponding to the number before correction for the first place, and if these do not match, It is determined that there is no corrected number corresponding to the number before correction.

  Also, before the correction (which matches the tens place value of the two digits of the original number OldNumTab of the note to be processed in the tablature 302 and the new number NewNumTab of the note to be processed in the tablature 302 ( It is determined whether or not the number after correction (before correction) associated with the number after correction matches the number of the first place of one digit. As a result of this determination, if they match, it is determined that there is a corrected number corresponding to the number before correction for the tens place, and if these do not match, It is determined that there is no corrected number corresponding to the number before correction.

  For example, the value of the original number OldNumTab of the note to be processed of the tablature 302 is 14, the value of the new number NewNumTab of the note to be processed of the tablature 302 is 2, and the table shows the number before correction. 1 and 2 are stored as corrected numbers. In this case, for the tens place, it is determined that there is a corrected number corresponding to the number before the correction, and for the first place, it is determined that there is no corrected number corresponding to the number before the correction. Is done.

  Next, in step S314, the CPU 1 determines whether there is a corrected number corresponding to the number before correction in a table stored in advance in the HDD 4 or the like at any digit. As a result of this determination, if there is a corrected number corresponding to the number before correction in any digit, the process of step S309 described above is performed. On the other hand, if there is no corrected number corresponding to the number before correction in any digit, the process proceeds to step S311 described above.

  As described above, in this embodiment, the difference between the note number of the note on the tablature 302 and the note number of the note on the staff 301 that is paired with the note on the tablature 302 is calculated. When the calculated note number is larger than the threshold NoteDiffThr, a value obtained by subtracting the tab staff tuning information 408 applied to the note of the tab staff 302 from the note number of the note of the staff 301 is a note of the tab staff 302. (See steps S301 and S302 in FIG. 10). Therefore, even if there is an error in the recognition of the tablature by pattern matching, the error can be corrected.

In this embodiment, the representative value of the difference between the note number of the note of the tablature 302 and the note number of the note of the staff 301 that is paired with the note of the tablature 302 is obtained for each string and measure. . Then, when bars having a representative value of the obtained note number difference larger than the threshold value are consecutive, an average of the representative values of the note number differences of the strings in the consecutive bars is obtained. If the variation in the average value of the representative values of the note numbers of the strings in the consecutive bars is within a predetermined range, the clef in the clef misrecognition area at the beginning of the consecutive bars Is added (see steps S19 to S22 in FIG. 6). Therefore, even if there is an error in the recognition of the staff notation by pattern matching, the error can be corrected.
As described above, in this embodiment, a score including a tablature can be recognized more accurately than before.

  In the present embodiment, the first representative value is calculated for each string from the difference between the note number of the note on the tablature 302 and the note number of the note on the staff 301 that is paired with the note on the tablature 302. , For each measure (see region 701a in FIG. 7A). Then, a second representative value is obtained for each string from the first representative value in a plurality of bars (for example, all bars) (see region 701b in FIG. 7A). If the second representative value is equal to or greater than the threshold Thr, the tab representative tuning information 408 of the corresponding string is changed using the second representative value (steps S203 to S205 in FIG. 9). See). Therefore, the tuning information can be applied to the tablature more appropriately than in the past.

  In the present embodiment, a measure having the same measure number as the measure number of the measure to which the note to be processed of the tablature 302 belongs is searched from the measures of the staff 301. Then, among the notes of the retrieved bar, the musical note 301 having the same positional relationship as the position of the note to be processed of the tab score 302 in the measure to which the note belongs is paired with the note of the tab score 302. It was made to detect as a note of (refer FIG. 8). Accordingly, even when the bar of the staff 301 and the bar of the tablature 302 are shifted in the horizontal direction, the note of the staff 301 that is paired with the note of the tablature 302 can be accurately detected.

  In this embodiment, in steps S20 to S22 in FIG. 6, the note number comparison table 702 created in step S19 is used to change the result of clef recognition in the staff 301. There is no need to do this. For example, when there are a plurality of continuous bars whose values stored in the region 702a are equal to or greater than the threshold value (see the portion displayed in gray in FIG. 7B), for those bars, step S23 in FIG. It is possible not to correct the tablature numbers in. In this way, it is possible to prevent the numbers in the tablature 302 from being inadvertently corrected when a temporary symbol or the like in pattern matching is erroneously recognized. Alternatively, the key signature other than the clef may be changed using the note number comparison table 702 created in step S19.

  In addition, the process allows the user to select whether or not to perform at least one of step S16 (FIG. 9), steps S20 to S22, and step S23 (FIG. 10) in FIG. May be executed. Further, in step S23 (FIG. 10), the content after the correction of the tablature numbers may be displayed, and when the correction is permitted by the user, the number of the corresponding tablature may be corrected.

  Further, steps S20 to S22 in FIG. 6 are not necessarily performed. Steps S20 to S22 may be performed before step S14. Furthermore, steps S305 to S310 and S312 to S314 in FIG. 10 are not necessarily performed.

  If the recognition result correction process in step S6 in FIG. 2 is performed after the combining process in step S5 as in this embodiment, the recognition result correction in step S6 in FIG. 2 is completed after the entire score is completed. Since the process is executed, it is preferable that the recognition result by pattern matching can be corrected more accurately, but this is not always necessary. For example, the recognition result correction process may be performed between step S4 and step S5.

The embodiment of the present invention described above can be realized by a computer executing a program. Further, a means for supplying the program to the computer, for example, a computer-readable recording medium such as a CD-ROM recording such a program, or a transmission medium for transmitting such a program may be applied as an embodiment of the present invention. it can. A program product such as a computer-readable recording medium in which the program is recorded can also be applied as an embodiment of the present invention. The above programs, computer-readable recording media, transmission media, and program products are included in the scope of the present invention.
In addition, each of the above-described embodiments is merely a specific example for carrying out the present invention, and the technical scope of the present invention should not be construed as being limited thereto. . That is, the present invention can be implemented in various forms without departing from the technical idea or the main features thereof.

It is a block diagram which shows embodiment of this invention and shows an example of a structure of the score recognition apparatus. It is a main flowchart which shows embodiment of this invention and demonstrates an example of a process of CPU. It is a figure which shows embodiment of this invention and shows an example of the score imported into a score recognition apparatus. It is a figure which shows embodiment of this invention and shows an example of a part template. It is a figure which shows embodiment of this invention and shows an example of tablature tuning information. It is a flowchart which shows embodiment of this invention and demonstrates an example of the recognition result correction process in step S6 of FIG. It is a figure which shows embodiment of this invention and shows an example of a note number comparison table. It is a flowchart which shows embodiment of this invention and demonstrates an example of the front part reference information acquisition process of FIG.6 S13. It is a flowchart which shows embodiment of this invention and demonstrates an example of the tuning information correction process in FIG.6 S16. It is a flowchart which shows embodiment of this invention and demonstrates an example of the tab score number correction process of step S23 of FIG.

Explanation of symbols

1 CPU
2 ROM
3 RAM
4 HDD
5 FDD
6 MIDI
7 Scanner Interface Circuit 8 Printer Interface Circuit 9 Keyboard Interface Circuit 10 CRT Interface Circuit 11 Scanner 12 Printer 13 Keyboard 14 CRT
100 Score recognition device

Claims (7)

Acquisition means for acquiring, from the image reading means, image information of a score including a tablature and a staff score paired with the tablature;
Recognizing means for recognizing which musical symbol the graphic belongs to by matching a graphic described in the score with a preset dictionary;
A calculating means for calculating a difference between a note number of the note of the tablature and a note number of the note of the staff notating the tablature note;
A musical score recognition apparatus comprising: a correction unit that corrects a music symbol recognized by the recognition unit based on a difference in note number calculated by the calculation unit. Search means for searching for a bar having the same bar number as the bar number of the bar to which the note of the tablature paired with the staff is paired, from the bars of the staff.
Detecting means for detecting, from the staff notation, notes having the same positional relationship as the position in the measure to which the note of the tab note belongs among the notes of the measure searched by the search means;
2. The musical score recognition apparatus according to claim 1, wherein the calculating means calculates a difference between a note number of the note of the tablature and a note number of the note detected by the detecting means.   The correcting means is based on the difference between the note number of the note on the tablature calculated by the calculation means and the note number of the note paired with the note on the tablature. The musical score recognition apparatus according to claim 1, wherein the number is corrected. Second calculation means for calculating a representative value of note number differences in a plurality of measures for each string of the tablature from the note number difference calculated by the calculation means;
4. The apparatus according to claim 1, further comprising: a second correction unit that corrects tuning information of a string in the tablature using the representative value calculated by the second calculation unit. The musical score recognition apparatus described in 1.   The correction means includes the difference between the note number of the note on the tablature calculated by the calculation means and the note number of the note paired with the note on the tablature, and the second correction means. 5. The score recognition apparatus according to claim 4, wherein the number of notes of the tablature is corrected based on the corrected tuning information.   The correction means, the difference between the note number of the note of the tablature and the note number of the note of the staff notated in a pair with the note of the tablature calculated by the calculation means is greater than or equal to a threshold value, 6. The musical score recognition apparatus according to claim 1, wherein a clef is added to a predetermined region at the beginning of a plurality of continuous measures in which variation of the difference is equal to or less than a threshold value. An acquisition step of acquiring image information of a score including a tablature and a notation that is paired with the tablature from the image reading unit;
A recognition step for recognizing which musical symbol the graphic belongs to by matching a graphic described in the score with a preset dictionary;
A calculation step of calculating a difference between a note number of the note of the tablature and a note number of the note of the staff notating the tablature note;
A computer program for causing a computer to execute a correction step of correcting a music symbol recognized by the recognition means based on a difference of note numbers calculated in the calculation step.

Images