A Fast Algorithm for Music Score Image Skew Detection
LIU Xiaoxiang,ZHANG Shusheng,WANG Jing,WANG Peng (Key Lab of Contemporary Design & Integrated Manufacturing Technology, Northwestern Polytechnical University, Xi'an 710072)
Optical music recognition (OMR) is an application of document image understanding to computer music. As a rule, an OMR system uses a scanned image of music score as input. In order to overcome the obstacles during next processes of staff-line detection and segment, skew detection of score image must be taken before recognition step is executed. In the paper, an approach is put forth to detect and correct skew score image. The algorithm is composed of two parts: the pre-processing of image and skew detection. The former using run-length coding of image removes the redundant pixels that are harmful to detect skew angle. The later employs horizontal projection of adjacent vertical image strips and cross-correlation algorithm to calculate the skew angle. The most advantage of the algorithm is that it takes very short processing time with high accuracy. Such improvement has been evaluated by the experimental results.
【CateGory Index】： TP391.4