This paper describes an automatic method for recognizing a blooming flower based on a photograph taken with a digital camera in natural scene. The problem of identifying an object against the background is known to be difficult. In this paper, we employ a photograph where the object (a blooming flower) is focused but the background is defocused. For extracting a flower region, we propose a new method that extracts a boundary by selecting a route with minimizing a sum of the local cost divided by the route length. Experiments were conducted for 600 pictures (20 pictures each for 30 species). A successful boundary extraction rate of 97% and a flower recognition rate of 90% were obtained.