In view of the problem of automatic recognition of opening-closing status for disconnecting switches in outdoor electrical equipment inspection in unattended substations, a method based on machine vision has thus been proposed for the opening-closing status recognition. The binary image of disconnecting switch can be obtained by dual threshold transformation, thus eliminating such unwanted targets as electrical insulators with lower gray value and sky background with higher gray value. The spatial filter has been employed to remove the interfering elements such as the transmission line and the supporting mechanism near the isolated switch areas, so as to extract the isolation switch. According to the number of connected regions projected in the horizontal and vertical directions of the binary images, a judgment can be made of the opening and closing status of the disconnecting switch. The experimental results show that this method can be used to recognize opening-closing status of disconnecting switches with a high accuracy under an outdoor complex background.