The attitude of an image sensing apparatus is detected, and optimum filter processing is selected based on the result of detection for extracting feature information of a subject, thus processing is simplified. While a shutter button is half-depressed, it is determined whether the image sensing apparatus is performing landscape image sensing or portrait image sensing based on attitude information obtained from an attitude sensor (S105). Then based on the result of determination, one of two filter processings in different scan directions is performed thereby a person's face in an image is detected. Then image sensing conditions at the next timing are determined based on the area of the person's face in the image. When the shutter button is full-depressed, image sensing is performed in accordance with the latest image sensing conditions, and an obtained image is stored into a storage medium such as a memory card.