The invention discloses a power equipment identification efficiency optimization method based on template tracking, and the method comprises the following steps: building a deep classification network, and carrying out the training of the deep classification network; performing video key frame detection and identification by using a deep classification network model; extracting a key frame targetidentification area as a target template, and extracting template features; carrying out template matching in a double range of a non-key frame target area, and calculating a matching rate; judging the relationship between the matching rate and a threshold value, if the extreme value of the matching rate is greater than or equal to the threshold value, updating the coordinates of the template region, re-extracting template features, and if the matching rate is less than the threshold value, skipping to a deep neural network model for identification; judging whether power equipment identification is ended or not, and if so, exiting. By using the method, the tracking speed is high, the invalid area recognized by the deep learning classifier is effectively reduced, and the recognition frame rate and the false detection rate are improved.