Song is a real occasion which is living with human life. Due to this, human had been using different accompanying instrument and discovering abstracts or features of song/music for better expression of their emotion (happiness, repentance and sadness). Even if, features of song /music have been set and standardized literally, but identifying in computerized way is somehow a challenging task. Among features of song/music scale is one sub-feature which play an important role in uniquely identifying melodies, since its properties are reflected to the melodic essence. The extraction and understanding of music scales are also essential in information analysis, retrieval and composition of music. In addition religious institutions like orthodoxy Christian, Muslim and including other institutions are very sensitive to music scales. Consequently, classic algorithms for identifying pentatonic scales have been designed based on the most popular scales; major and minor scales. In this research, we designed a model for identifying well-known Ethiopian traditional lyre Bagana song scales, such as Selamta, Wanen, Chernet and Bati Major. To identify the scales primarily we have compare three sample pitch detection approaches. Approaches like time, frequency and hybrid domain such as PYIN, ACF, CCF, SHS, and CAF. When we apply the detector algorithm on 713 melodies which were prepared for pitch detection we got the accuracy 98.4, 74%, 76%, 68% and 72% respectively. In this study, rule base approach is used by using PYIN, we extracted pitches frequency and it is converted to pitch note. Based on pitch occurrence ranking is performed. When extra pitch is not occurred normal identification were going to be performed, otherwise, by setting threshold value identification is done. Finally, the proposed model is tested on the data collected about 1357 of different audio category from concerned Bagana training institutes. The experimental result is shown (accuracy for single and double strumming 94.7% and 81.65% respectively).