Terrain Feature Extraction Method of Digital Transmission Line based on Depth Self Encoder

A terrain feature extraction method of digital transmission line based on depth self encoder is proposed. The proposed method adopts maximum likelihood estimation algorithm and minimum mean square error (MMSE) method, which can extract terrain features with high accuracy. The technology has been tested in three different cases, including case 1, where there is no obstacle in front of the transmitter, case 2, where there is an obstacle in front of the transmitter, and case 3, where there is an obstacle behind the receiver. The results show that for all three cases, features can be extracted with reasonable accuracy. This technique uses the depth information of a single channel to determine the height of a hill or valley in an image captured at different angles relative to the horizontal plane. The height is estimated by comparing the distance between points on two images taken at different angles and intervals