This paper propose a phone boundary detection method based on the phonological attributes posterior probability, and taking these features and boundary information in Conditional Random Fields based phone recognition system. Firstly, calculate the angles between posterior probability vectors of adjacent frames, and then mark the frames with maximum angle as the boundary candidates. Secondly, remove the false phone boundaries through several restrictions in these boundary candidates. Finally, the combination of phonological attributes and phone boundaries are presented as the observation vectors of Conditional Random Fields, experimental results show that the accuracy rate of phoneme recognition is superior to the base system which use phonological attributes features only.