Electronic image capture with reduced noise

A method of choosing multiple exposures for capturing an image of a scene with reduced noise, comprising:

• providing to an image processing apparatus image data representative of a scene;
• based on the image data and a noise model that characterizes the noise performance of the image capture system, identifying one or more portions of the scene that are predicted to contribute disproportionately to the visible noise in said image formed from said image data;
• and
• based on the identified portions of the scene and the noise model, determining an exposure pattern for the image capture system that is predicted to produce multiple exposures of the scene that are combinable to produce an image with a minimal predicted noise.

A method of capturing a reduced noise image of a scene using an electronic image capture system including an image processing apparatus, the method comprising the steps of providing to the image processing apparatus image data representing at least one monitored image, and using the image processing apparatus to:

• i) generate from the image data a tonal importance mapping;
• ii) use the tonal importance mapping to identify one or more portions of the scene that would contribute disproportionately to the noise present in an image derived from said image data;
• iii) choose a multiple exposure pattern including at least one exposure to reduce the noise contribution from at least one of said portions;
• and
• iv) use said desired multiple exposure pattern to capture multiple images of said scene and combine said multiple images to produce the reduced noise image.