The purpose of the present invention is to perform pulse wave detection that is robust against disturbance factors. A pulse wave detection device 1 color converts a frame image of a moving image from RBG components to HSV components, and identifies a skin section using the skin color of a user prepared in advance with an H component. Next, the pulse wave detection device 1 converts the skin section of the frame image to YIQ components, and takes Qm as a pulse wave signal, Qm being obtained by averaging the Q values of the pixels. After carrying out the abovementioned processing on each frame image, the pulse wave detection device 1 obtains a chronological change in the pulse wave signal Qm and outputs said change as a pulse wave. As a result, it is possible to exclude disturbance factors such as the background from the pulse wave detection target, and successfully detect a pulse wave. Furthermore, the identification of the skin section is performed by using the H component to identify candidate pixels, and using the S component to narrow down the candidate pixels to the target, such that high-precision identification of the skin section is made possible.
