Method and apparatus for determining summation of pixel characteristics for rectangular region of digital image avoiding non-aligned loads using multiple copies of input data

A method comprising:

• receiving a data set;
• storing the data set in a first buffer;
• storing the data set in a second buffer at an offset relative to the first buffer;
• determining whether to load a first data element of the data set from the first buffer or the second buffer based on a word alignment of the first data element in the first buffer and in the second buffer, wherein the first data element is a corner of a first region;
• based on the determining, loading the first data element from either the first buffer or the second buffer, wherein the loading further retrieves a second data element that is a corner of a second region;
• performing a calculation on the first region using the first data element;
• and
• performing a calculation on the second region using the second data element.

A device comprising:

• a first buffer;
• a second buffer;
• at least one processor coupled to the first buffer and the second buffer;
• and
• a non-transitory computer readable storage medium storing a program for execution by the at least one processor, the program including instructions to:
• receive an image;
• store a representation of the image in the first buffer;
• store the representation of the image in the second buffer at an address offset relative to the representation in the first buffer;
• perform a feature calculation on a first region of the image and on a second region of the image that overlaps the first region by:
• determining whether to retrieve a first data element from the first buffer or the second buffer based on a word alignment of the first data element, wherein the first data element is within the first region;
• and
• in response to the word alignment of the first data element, retrieving the first data element and a second data element from one of the first buffer or the second buffer, wherein the second data element is within the second region;
• performing the feature calculation on the first region using the first data element;
• and
• performing the feature calculation on the second region using the second data element.