Low complexity auto-exposure control for computer vision and imaging systems

An apparatus for performing automatic exposure control (AEC) comprising:

• an image sensor for capturing a first digital image of a scene;
• at least one processor or circuitry coupled to the image sensor and configured to;
• apply a first set of values to one or more exposure control parameters while the first digital image is captured;
• perform object detection on the first digital image, wherein the object detection identifies one or more objects in the first digital image, wherein the object detection outputs a first set of features extracted from the first digital image, wherein a feature is a numerical value indicating gradients between a set of neighboring pixels, and wherein the first set of features includes a count of occurrences in the first digital image of each feature of a set of possible features;
• obtain a mask associated with one or more objects in the first digital image, the mask having a zero value or a non-zero value for each feature of the set of possible features, wherein the mask has a non-zero value for each feature of the set of possible features that is used by the object detection to classify contents of an image as the one or more objects;
• apply the mask to the first set of features to determine a subset of the first set of features, the subset of the first set of features including each feature from the first set of features that corresponds to a non-zero value in the mask;
• compute a first score using the subset of the first set of features;
• and
• generate, based on the first score, an updated set of values for applying to the one or more exposure control parameters for capturing a subsequent digital image of the scene.

A non-transitory computer readable medium having instructions stored thereon that, when carried out by one or more processors, causes the one or more processors to:

• capture a first digital image of a scene using an image sensor while applying a first set of values to one or more exposure control parameters;
• perform object detection on the first digital image, wherein the object detection identifies one or more objects in the first digital image, wherein the object detection outputs a first set of features extracted from the first digital image, wherein a feature is a numerical value indicating gradients between a set of neighboring pixels, and wherein the first set of features includes a count of occurrences in the first digital image of each feature of a set of possible features;
• obtain a mask associated with one or more objects in the first digital image, the mask having a zero value or a non-zero value for each feature of the set of possible features, wherein the mask has a non-zero value for each feature of the set of possible features that is used by the object detection to classify contents of an image as the one or more objects;
• apply the mask to the first set of features to determine a subset of the first set of features, the subset of the first set of features including each feature from the first set of features that corresponds to a non-zero value in the mask;
• compute a first score using the subset of the first set of features;
• and
• generate, based on the first score, an updated set of values for applying to the one or more exposure control parameters for capturing a subsequent digital image of the scene.

A system for performing automatic exposure control (AEC) comprising:

• means for capturing a first digital image of a scene using an image sensor while applying a first set of values to one or more exposure control parameters;
• means for performing object detection on the first digital image, wherein the object detection identifies one or more objects in the first digital image, wherein the object detection outputs a first set of features extracted from the first digital image, wherein a feature is a numerical value indicating gradients between a set of neighboring pixels, and wherein the first set of features includes a count of occurrences in the first digital image of each feature of a set of possible features;
• means for obtaining a mask associated with one or more objects in the first digital image, the mask having a zero value or a non-zero value for each feature of the set of possible features, wherein the mask has a non-zero value for each feature of the set of possible features that is used by the object detection to classify contents of an image as the one or more objects;
• means for applying the mask to the first set of features to determine a subset of the first set of features, the subset of the first set of features including each feature from the first set of features that corresponds to a non-zero value in the mask;
• means for computing a first score using the subset of the first set of features;
• and
• means for generating, based on the first score, an updated set of values for applying to the one or more exposure control parameters for capturing a subsequent digital image of the scene.