Hidden background cameras used to realize Diminished Reality (DR), which allows the observer to see through obstacles, are installed behind the obstacles. In order to realize DR, it is necessary to measure the positions of these cameras with respect to the observer’s position, but this is difficult because the cameras cannot be seen directly from the observer’s position. The objective of this study is to develop a method for measuring the poses of hidden background cameras through obstacles using direction finding of radio sources. Two calibration methods, the intersection method and the relative pose constraint method, were implemented and evaluated for this purpose. The intersection method measures the direction from several different points to the hidden background cameras and estimate the 3D position of the camera as the intersection of the measured directions. The relative pose constraint method uses the relative pose between hidden background cameras measured in advance as a constraint to estimate the 3D poses of cameras. The results showed that the relative pose constraint method was more accurate than the intersection method, and able to calibrate the camera with a position error of less than 100 mm.