The invention discloses an intelligent truck crane video identification method for a power transmission and distribution line channel. The method is characterized by collecting video images through a front-end camera and carrying out decompression on the collected video images to enable the video images to be decomposed into YUV pictures required for intelligent analysis; converting the YUV pictures into RGB pictures; decomposing the pictures into 25 frames of RGB pictures within a second, and randomly drawing 10 frames of pictures from the RGB pictures to carry out analysis and comparison; comparing the next frame of picture with the previous frame of picture, updating background environment in real time by utilizing self-learning, and drawing a truck crane therefrom; and when locking to a target object, carrying out gray level calculation on a large-size dangerous vehicle, and after identification and determination of the specific shape of the truck crane, giving an alarm. The method can carry out identification on truck crane conditions of the power transmission and distribution line channel conveniently and timely.