The invention provides a single-camera ranging method. The single-camera ranging method comprises the following steps: S1, pasting a target image in front of a sampled object, and placing and starting a camera; S2, converting image information collected by the camera into a black-white image; S3, sampling the black-white image by using a micro processor, wherein M*N sampling points are taken; S4, defining the sampling points in a black image area as feature points, and calculating the average number of feature points in each row; S5, changing a relative distance between the camera and the sampled object, and repeating the steps S1 to S4; S6, establishing a database of a relationship between the relative distance and the average number of the feature points in each row, and obtaining a nonlinear approximate equation for the relative distance and the average number of the feature points in each row by using an MATLAB tool; S7, pasting the target image on the surface of the sampled object, calculating the average number of the feature points in each row in the black-white image generated by the sampled object, and substituting into the nonlinear approximate equation for the relative distance and the average number of the feature points in each row to obtain the relative distance between the sampled object and the camera.