With different imaging mechanisms of infrared light and visible light， the image information could not be synchronized， and the dual spectral system of infrared light and visible light could not be used simultaneously to achieve the ranging function. In this paper， the boundary of the target to be measured was used as a common feature， infrared and visible light cameras were used to collect the common features of the same target， the collected image was smoothed by the Gaussian operator， and then the Laplace expansion operator was used to detect the boundary features， linear fitting was performed on the pixel position data of the target boundary at different distances， and a ranging error model was established， so as the boundary feature matching algorithm was generated. The research results showed that the boundary feature matching algorithm could enable the infrared visible binocular system to overcome the problem of different imaging mechanisms， and achieve ranging. The ranging range could reach 40 m， and the maximum relative error is 0.975%.