A visible image of gas discharge contains rich information. This study investigates the visible light information of surface microdischarge (SMD) under different operation modes and develops a method to realize online monitoring of the operation modes based on a gray level histogram (GLH). Two SMD sources were used to verify the effectiveness of the method. The distribution curve in the GLH gradually shifted to a large gray level as the operation mode transferred from ozone mode to NO x mode. The mean, variance, energy, entropy, and uniformity threshold values used to distinguish the operation modes were found to be approximately the same in both sources. Finally, a simplified distribution map based only on entropy and uniformity is proposed for the fast operation mode recognition of SMD.