The process of optical music recognition, converting a music sheet into notes and pitches in a digital representation, resembles the process of optical character recognition. It involves a lot of steps from the broad spectrum of image processing. Optical music recognition includes the steps from pre-processing the input image, to classifying extracted musical features into the output that is a string representation of the notes.
DDCBBAGGFEDDnDEDDAFEDGBDnDCBBAGGFEDDnCBABAGAEFG
Images | ref/out | Label | Fail count | Fail rate |
im1s.jpg | ref | DDCBBAGGFEDDn DEDDAFEDGBDn DCBBAGGFEDDn CBABAGAEFG |
0 | 0% |
out | DDCBBAGGFEDDn DEDDAFEDGBDn DCBBAGGFEDDn CBABAGAEFG |
|||
im2s.jpg | ref | AAFABAFFEFEn AAFABAFEFEn DDFABBDBn AAFABAFEFE |
3 | 7% |
out | AAFABAFFEFEn AAFABAFEFEn CDFABBDBn GGFABAFEFE |
|||
im3s.jpg | ref | ACBADDDEBCn AAACBAGGFEDCBAn ACBADDDEBCn AAACBAGDABG |
1 | 2% |
out | ACBADDDEBCn DAAACBAGGFEDCBAn ACBADDDEBCn AAACBAGDABG |
|||
im4s.jpg | ref | GGEEGGDDEFGABGGn GEEGGDDDCDEADGn EEDCCBBCDBAGCCn CAACCGGGACGDC |
1 | 2% |
out | GGEEGGDDEFGABGGn GEEGGDDDCDEADGn EEDCCBBCDBAGCCn CAACCGGACGDC |
|||
im5s.jpg | ref | GGABGBADGGABGFn GGABCBAGFDEFGGn EFEDEFGDEDCDn EFEDEFGEDGFAGG |
1 | 2% |
out | GABGBADGGABGFn GGABCBAGFDEFGGn EFEDEFGDEDCDn DEFEDEFGEDGFAGG |
|||
im6s.jpg | ref | CCGGAAGFFEEDDCn GGFFEEDGGFFEEDn CCGGAAGFFEEDDC |
1 | 2% |
out | CCGGAAGFFEEDCn GGFFEEDGGFFEEDn CCGGAAGFFEEDDC |
|||
Result | Total fail count | Success rate | ||
7 | 97,7% |