Contrast detection autofocus using multi-filter processing and adaptive step size selection
AssigneeIntel Corporation
InventorEvgeny Krestyannikov
Life of this patent
Estimating the term…
Abstract
Techniques related to contract detection autofocus for imaging devices and, in particular, to contrast detection autofocus using multiple filters to generate adaptive step sizes are discussed. Such techniques may include determining a ratio of filter responses for filters applied to a portion of a scene, generating a distance from focus based on the ratio, and determining a lens position movement based on an adaptive step size generated using the distance from focus and a confidence level associated with the distance from focus.
Claims20
A method for providing contrast detection autofocus for an imaging device comprising:
- determining, for a region of interest of a scene, a ratio of a first response from a first filter applied to the region of interest and a second response from a second filter applied to the region of interest, wherein the first filter and the second filter comprise band pass filters, the first filter has a wider pass band than the second filter, and the first and second filters are both applied only to image data associated with the region of interest;
- generating a distance from focus and a confidence level associated with the distance from focus, wherein the distance from focus is based at least in part on the ratio;
- determining a lens position movement based at least in part on the distance from focus and the confidence level, wherein determining the lens position movement comprises adaptively generating a multiplier based on the distance from focus, the confidence level, and a current lens position and determining a product of the multiplier and a reference step size;
- and
- focusing the imaging device based at least in part on the lens position movement.
A system for providing contrast detection autofocus for an imaging device, comprising:
- a memory configured to store image data;
- and
- a central processing unit coupled to the memory, the central processing unit to:
- determine, for a region of interest of a scene, a ratio of a first response from a first filter applied to the region of interest and a second response from a second filter applied to the region of interest, wherein the first filter and the second filter comprise band pass filters, the first filter has a wider pass band than the second filter, and the first and second filters are both applied only to image data associated with the region of interest;
- generate a distance from focus and a confidence level associated with the distance from focus, wherein the distance from focus is based at least in part on the ratio;
- and
- determine a lens position movement based at least in part on the distance from focus and the confidence level, wherein to determine the lens position movement, the central processing unit is to adaptively generate a multiplier based on the distance from focus, the confidence level, and a current lens position and determine a product of the multiplier and a reference step size, and focus the imaging device based at least in part on the lens position movement.
At least one non-transitory machine readable medium comprising a plurality of instructions that in response to being executed on an imaging device, cause the imaging device to provide contrast detection autofocus by:
- determining, for a region of interest of a scene, a ratio of a first response from a first filter applied to the region of interest and a second response from a second filter applied to the region of interest, wherein the first filter and the second filter comprise band pass filters, the first filter has a wider pass band than the second filter, and the first and second filters are both applied only to image data associated with the region of interest;
- generating a distance from focus and a confidence level associated with the distance from focus, wherein the distance from focus is based at least in part on the ratio;
- determining a lens position movement based at least in part on the distance from focus and the confidence level, wherein determining the lens position movement comprises adaptively generating a multiplier based on the distance from focus, the confidence level, and a current lens position and determining a product of the multiplier and a reference step size;
- and
- focusing the imaging device based at least in part on the lens position movement.
Dependencies are inferred from the claim text.