JPS5957384A - Recognition of musical notation - Google Patents
Recognition of musical notationInfo
- Publication number
- JPS5957384A JPS5957384A JP57168153A JP16815382A JPS5957384A JP S5957384 A JPS5957384 A JP S5957384A JP 57168153 A JP57168153 A JP 57168153A JP 16815382 A JP16815382 A JP 16815382A JP S5957384 A JPS5957384 A JP S5957384A
- Authority
- JP
- Japan
- Prior art keywords
- dictionary
- matching
- notation
- data
- recognition
- Prior art date
- Legal status (The legal status is an assumption and is not a legal conclusion. Google has not performed a legal analysis and makes no representation as to the accuracy of the status listed.)
- Granted
Links
Classifications
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06V—IMAGE OR VIDEO RECOGNITION OR UNDERSTANDING
- G06V10/00—Arrangements for image or video recognition or understanding
- G06V10/70—Arrangements for image or video recognition or understanding using pattern recognition or machine learning
- G06V10/74—Image or video pattern matching; Proximity measures in feature spaces
- G06V10/75—Organisation of the matching processes, e.g. simultaneous or sequential comparisons of image or video features; Coarse-fine approaches, e.g. multi-scale approaches; using context analysis; Selection of dictionaries
- G06V10/751—Comparing pixel values or logical combinations thereof, or feature values having positional relevance, e.g. template matching
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06V—IMAGE OR VIDEO RECOGNITION OR UNDERSTANDING
- G06V30/00—Character recognition; Recognising digital ink; Document-oriented image-based pattern recognition
- G06V30/10—Character recognition
- G06V30/30—Character recognition based on the type of data
- G06V30/304—Music notations
Abstract
Description
【発明の詳細な説明】
この発明は、楽譜の1,1llI等の記号を2次元マツ
チング法で認識するための方法に関する。DETAILED DESCRIPTION OF THE INVENTION The present invention relates to a method for recognizing symbols such as 1 and 1llI in a musical score using a two-dimensional matching method.
この発明では、第1図に示すような五線等を消去した画
像エリア1からマツチングを行なうための対象物(この
例ではr>を切出イための切出し範囲2を定め、切出し
範囲2で定められた切出し画像をマツチングを行なうエ
リア3に移動する。In this invention, a cropping range 2 is defined for cropping an object for matching (in this example, r The determined cutout image is moved to area 3 where matching is to be performed.
なお、このマツチングエリア3Ii切出し画像よりも少
し大きくなっている。そして、別に予め用意しておいた
マツチング用の辞書を、マツチングエリア3内で動かし
て両者のマツチングを行なうよ 。Note that this matching area 3Ii is slightly larger than the cutout image. Then, a matching dictionary prepared separately in advance is moved within the matching area 3 to match the two.
うにするのであ7I0
ここにおいて、第2図(A)に示す画像(#)を、辞書
に登録されている同図(11)に示すような記号(埼)
とマツチングを行なう場合、その形状の異なる部分はP
及びQにおいて顕著である。そして、両津に関しては第
3図(A)に示すように画像部分を11“とし、その他
の領域を10”と−4−^と共に、辞書には第3図中)
に示すような肖像部分を% 1 Nとして周囲を全て%
□lとするデータと、同図(C) K示すように画像部
分を11′として周囲を全て1−1#とするデータとを
用意しておき、第2[1VIAlのl’ if!S分と
同図(I3)のQ ?4U分とを陥理演11により比較
するようにしている。すなわち、第3図(A)に示すP
l、 P2 と同図(13)に示″tQs 、 Q3及
び同図((、’)K示すQ2 、 Q4との画素データ
毎の論理積を求め、後述する類似度からその積値を減算
するようにして類似程度を判断するようにしていイ)。7I0 Here, the image (#) shown in Fig. 2 (A) is replaced with the symbol (Sait) shown in Fig. 2 (11) registered in the dictionary.
When matching is performed, the parts with different shapes are P
and Q. Regarding Ryotsu, as shown in Figure 3 (A), the image part is set to 11", and the other areas are 10" and -4-^ (in Figure 3) in the dictionary.
The portrait part as shown in is % 1 N, and the surrounding area is %
□Prepare the data for l' and the data for the image part to be 11' and the surrounding area to be 1-1# as shown in Figure (C) K. S minute and Q in the same figure (I3)? I am trying to compare it with the 4U portion using trick theory 11. That is, P shown in FIG. 3(A)
1, P2 and "tQs, Q3 shown in (13) in the same figure, and Q2, Q4 shown in ((,')K in the same figure for each pixel data are calculated, and the product value is subtracted from the degree of similarity described later. The degree of similarity is determined in this way.
つまり、第3図囚と(B)との場合には、PI XQI
+P2 xQa=0であるから類イυ度に影響はない
。これに対して、第3図(A)と((こ)との場合にi
、’r、 、 PI XQ2 +P2 XQ4=−6
であるから類似度から[6Jを減算し、その減算値を類
似度とする。したがって、前者の場合には画像が辞得よ
り太き(ても知什j度には変化がな(、形状の違いが評
価されない。これに対し、後者の場合には画像が大きい
とその分だけ類似度から減算されるので、辞劉との違い
が強調されて評価されることにt「る。In other words, in the case of the prisoner in Figure 3 and (B), PI
Since +P2 xQa=0, there is no effect on the degree of similarity. On the other hand, in the case of Fig. 3(A) and (((this)), i
, 'r, , PI XQ2 +P2 XQ4=-6
Therefore, [6J is subtracted from the similarity, and the subtracted value is taken as the similarity. Therefore, in the former case, the image is thicker than the size (even if there is no change in the degree of familiarity), the difference in shape is not evaluated.On the other hand, in the latter case, if the image is larger, Since the difference from Ji Liu is subtracted from the similarity, the difference from Ji Liu will be emphasized and evaluated.
ここで、第3図(A)の画像をgとし、第3図(均に示
すような′1′及び10″のデータの記憶で成る辞書な
fとし、同図((−’)に示すような11′′及び′−
1′ のデータで成る辞書をf′とした場合、fXgに
よる汁のマツチング方式では、丑と科で高い類似度を得
、竹と1では指状の差力槽わ」する。Here, the image in Fig. 3 (A) is g, the dictionary consisting of the memory of data ``1'' and 10'' as shown in Fig. 11'' and '-
If f' is a dictionary consisting of data of 1', then in the soup matching method using fXg, a high degree of similarity is obtained between the ox and the family, and a finger-like differential force is obtained between the bamboo and 1.
これにより吐と明の区別ができろ。これに対t7、f’
Xgのマツチング方式では、件と壮において画像gのず
れに敏感とfcす、類イリ度は低く、科と町の類似度と
同じになり、址と1の区別が[7にくい。This allows you to distinguish between tō and ming. On the other hand, t7, f'
In the matching method of Xg, it is sensitive to the deviation of the image g in case and place, and the degree of similarity is low, the degree of similarity is the same as that of family and town, and it is difficult to distinguish between place and place.
また、−のマツチングにおいて、fXgのマツチング方
式では―と什で−は壮に重なってしまい、類似度が高(
なり、1と−の類似度と同じになり、せと−の区別がで
きない。こJlに対して、f’Xgのマツチング方式に
おいては、―と丑で形状の違いが強調されてd、、c4
Wり度が小さくなるので、#と時の区別ができるので、
tI)る。In addition, when matching -, in fXg's matching method, - and - overlap significantly, and the degree of similarity is high (
The degree of similarity is the same as that between 1 and -, and it is impossible to distinguish between set and -. For this Jl, in the matching method of f'Xg, the difference in shape is emphasized between - and ox, and d, , c4
Since the degree of W becomes smaller, it is possible to distinguish between # and hour.
tI).
この発明はこのような観点からなさねたものであり、辞
書にゝゝIN及び10′′のデータと、′1“及び%
1#のデータとを予め2種類用意しておき、画像デー
タと辞書のデータとを比較すること11Cより、セと目
との誤判断がないような記号の認識が可能となる。This invention was made from this point of view, and the dictionary contains the data ``IN'' and 10'', and the data ``1'' and %.
By preparing two types of data 1# in advance and comparing the image data with the dictionary data 11C, symbols can be recognized without misjudgment between the center and the eyes.
次に、この発明による類似贋のlq方力法)呪制する。Next, the method of similar counterfeiting according to this invention is controlled.
第4図に示すように対象画像gの黒の部分を11“とし
、白のglt(Jヲ’ o ’ トL、黒(% 1 N
>の総計を1とすると、h[!号沖についてはfXg
=b 曲曲曲(1)の辞書を用い、類似度
Sfgを
Sfg=7−X 100 [%コ 川・曲間
(2)で求める。そして、記号妙以外の記号、たとえば
す、H等の楽譜Hr4号に関しては
f’Xg=c 曲曲曲(3)を用い、その類
イ1′)度Sfgを
Sfg = −X IO(+ [%] 曲曲曲(4)
皿
なる計算で求めるようにしている。なお、上記画像と辞
書とのマツチングは構成画累毎に行なうようにしている
ので、結局上記(1)及び(3)式は画素データをG(
i、j)及びF(’ e J )、F”(+。As shown in FIG.
> is 1, then h[! fXg for No.Oki
=b Using the dictionary of the song (1), the similarity Sfg is calculated as Sfg=7-X 100
Find it using (2). Then, for symbols other than the symbol, such as Su, H, etc., f'Xg=c Using the song (3), the degree Sfg of that kind is Sfg = -X IO(+ [ %] Song (4)
I try to find it using basic calculations. Furthermore, since the matching between the above image and the dictionary is performed for each constituent picture, the above equations (1) and (3) ultimately convert the pixel data into G(
i, j) and F(' e J ), F''(+.
j)とすると、次の式のようになる。j), the following formula is obtained.
fXg41” (i r j)・(i(i 、 j )
=== (5)目
f’Xg=↓゛ΣF’(i、j)・G(i、j)・曲・
(6)1」
以上のようKこの発明の認識方法にょり、ば、ヰとヰ以
外の記号に対してそねぞ」1異なるデータで成る辞書を
用意しておき、片と−の認識の困難性を除去しているの
で、2次元マツチング法によっても確実な記号の認識を
行なうことが可能となる。fXg41” (i r j)・(i(i, j)
=== (5)th f'Xg=↓゛ΣF'(i, j)・G(i,j)・song・
(6) 1 As described above, with the recognition method of this invention, it is possible to recognize symbols other than ba, ヰ and ヰ. Since this difficulty is removed, it is possible to reliably recognize symbols even with the two-dimensional matching method.
第1図は画像エリアからマツチングエリアな切出す様子
を説明するための図、第2図(A)及び(I3)はこの
発明における片と−の認識の様子を説明するための図、
第3図(A)〜(C1は画像データと辞書データとを説
明するための図、第4図はこの発明による% 1 #
、 % 0 #の振り分を示す図である。
1・・・画像エリア、2・・・切出し範囲、3・・・マ
ツチングエリア、P、Q・・・マツチング対象。FIG. 1 is a diagram for explaining how a matching area is cut out from an image area, and FIGS. 2 (A) and (I3) are diagrams for explaining how pieces and - are recognized in this invention.
FIGS. 3(A) to (C1 are diagrams for explaining image data and dictionary data, and FIG. 4 is a diagram for explaining image data and dictionary data.
, % 0 # is a diagram showing the allocation. 1... Image area, 2... Cropping range, 3... Matching area, P, Q... Matching target.
Claims (1)
、対象記号に対して11N及びゝ−1′のデータで成る
辞書データfと、′1#及び′ONのデータで成る辞書
データf′とを用意しておき、認識対象領域からマツチ
ングの対象記号を% 11及び′□Iのデータgで抽出
すると共に、fXg=b又はf’Xg=cを求め、前記
認識対象領域のなる式で類似度を求めるようKしたこと
を特徴とする楽譜記号の認識方法。In a method for recognizing music score symbols using a two-dimensional matching method, dictionary data f consisting of data of 11N and -1' and dictionary data f' consisting of data of '1# and 'ON' are prepared for the target symbol. Then, extract the target symbol for matching from the recognition target area using the data g of %11 and '□I, find fXg=b or f'Xg=c, and calculate the similarity using the formula for the recognition target area. A method for recognizing music score symbols, characterized by the fact that K is used to ask for information.
Priority Applications (1)
Application Number | Priority Date | Filing Date | Title |
---|---|---|---|
JP57168153A JPS5957384A (en) | 1982-09-27 | 1982-09-27 | Recognition of musical notation |
Applications Claiming Priority (1)
Application Number | Priority Date | Filing Date | Title |
---|---|---|---|
JP57168153A JPS5957384A (en) | 1982-09-27 | 1982-09-27 | Recognition of musical notation |
Publications (2)
Publication Number | Publication Date |
---|---|
JPS5957384A true JPS5957384A (en) | 1984-04-02 |
JPH0259509B2 JPH0259509B2 (en) | 1990-12-12 |
Family
ID=15862795
Family Applications (1)
Application Number | Title | Priority Date | Filing Date |
---|---|---|---|
JP57168153A Granted JPS5957384A (en) | 1982-09-27 | 1982-09-27 | Recognition of musical notation |
Country Status (1)
Country | Link |
---|---|
JP (1) | JPS5957384A (en) |
Cited By (2)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
EP0278522A2 (en) * | 1987-02-12 | 1988-08-17 | Kabushiki Kaisha Toshiba | Microprocessor |
US5864631A (en) * | 1992-08-03 | 1999-01-26 | Yamaha Corporation | Method and apparatus for musical score recognition with quick processing of image data |
-
1982
- 1982-09-27 JP JP57168153A patent/JPS5957384A/en active Granted
Cited By (2)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
EP0278522A2 (en) * | 1987-02-12 | 1988-08-17 | Kabushiki Kaisha Toshiba | Microprocessor |
US5864631A (en) * | 1992-08-03 | 1999-01-26 | Yamaha Corporation | Method and apparatus for musical score recognition with quick processing of image data |
Also Published As
Publication number | Publication date |
---|---|
JPH0259509B2 (en) | 1990-12-12 |
Similar Documents
Publication | Publication Date | Title |
---|---|---|
Hu et al. | Wordsup: Exploiting word annotations for character based text detection | |
US8234277B2 (en) | Image-based retrieval for high quality visual or acoustic rendering | |
JP2000339477A (en) | Method for expressing object, method for searching object, device for expressing or searching object, computer program, computer system and computer readable storage medium | |
EP0360267A3 (en) | Processor for process control | |
JP2004054957A (en) | Method of representing data distribution, method of representing data element, descriptor of data element, method of collating or classifying query data element, device set to execute the method, computer program, and computer readable storage medium | |
Vo et al. | Deep learning for vietnamese sign language recognition in video sequence | |
JPWO2013147022A1 (en) | Memory having set operation function and set operation processing method using the same | |
JPS5957384A (en) | Recognition of musical notation | |
KR100406508B1 (en) | System and method for processing color image | |
Roth | An approach to recognition of printed music | |
JP3327127B2 (en) | Image presentation device | |
Rad et al. | Digital image forgery detection by edge analysis | |
JPS60160483A (en) | Recognizing device of character | |
JP2985243B2 (en) | Character recognition method | |
Ip et al. | Affine invariant retrieval of shapes based on hand-drawn sketches | |
JPH0476159B2 (en) | ||
JPS58121483A (en) | Picture processing method | |
JPH0922442A (en) | Electronic management system for image document data | |
JPS63257073A (en) | Approximate graphic generating device | |
Tunyasanon et al. | Musical chords transposer for captured image based on Optical Character Recognition | |
CN113569871A (en) | Library automatic book-making method and system based on deep learning | |
CN117197561A (en) | Method for identifying year of coin | |
JPH0351982A (en) | Character string recognition system | |
Ibrahim | Robust Multiscale Shape Representation and Description | |
Aanæs et al. | Factorization with erroneous data |