Methods and apparatus to perform remote monitoring

A system to select areas of an image for object detection, the system comprising:

• at least one memory;
• machine readable data;
• and
• a field programmable logic device configurable by the machine readable data to at least:
• subtract first image data from second image data to determine a difference image;
• convert the difference image into a modified image, respective pixels of the modified image to have one of a plurality of pixel values including a first pixel value and a second pixel value;
• determine respective proportions for respective multi-pixel image blocks of the modified image, the multi-pixel image blocks respectively including multiple pixels of the modified image, the proportions representative of quantities of pixels having the first pixel value in the corresponding multi-pixel image blocks of the modified image;
• select first ones of the multi-pixel image blocks based on the respective proportions, the first ones of the multi-pixel image blocks to be processed to detect one or more objects;
• and
• skip processing of second ones of the multi-pixel image blocks that are not selected based on the respective proportions.

At least one non-transitory computer readable medium comprising computer readable instructions to cause one or more processors to at least:

• subtract first image data from second image data to determine a difference image;
• convert the difference image into a modified image, respective pixels of the modified image to have one of a plurality of pixel values including a first pixel value and a second pixel value;
• determine respective proportions for respective multi-pixel image blocks of the modified image, the respective multi-pixel image blocks respectively including multiple pixels of the modified image, the respective proportions representative of quantities of pixels having the first pixel value in the respective multi-pixel image blocks of the modified image;
• select first ones of the multi-pixel image blocks based on the respective proportions, the first ones of the multi-pixel image blocks to be processed to detect one or more objects;
• and
• skip processing of second ones of the multi-pixel image blocks that are not selected based on the respective proportions.

A method to select areas of an image for object detection, the method comprising:

• subtracting first image data from second image data to determine a difference image;
• converting the difference image into a modified image, respective pixels of the modified image to have one of a plurality of pixel values including a first pixel value and a second pixel value;
• determining respective percentages for respective multi-pixel image blocks of the modified image, the respective multi-pixel image blocks respectively including multiple pixels of the modified image, the respective percentages representative of quantities of pixels having the first pixel value in the respective multi-pixel image blocks of the modified image;
• selecting first ones of the multi-pixel image blocks based on the respective percentages, the first ones of the multi-pixel image blocks to be processed to detect one or more objects;
• and
• skipping processing of second ones of the multi-pixel image blocks that are not selected based on the respective percentages.

An apparatus to select areas of an image for object detection, the apparatus comprising:

• memory means;
• instructions;
• and
• processing means, the processing means to be controlled by the instructions, the processing means to:
• subtract first image data from second image data to determine a difference image;
• convert the difference image into a modified image, respective pixels of the modified image to have one of a plurality of pixel values including a first pixel value and a second pixel value;
• determine respective proportions for respective multi-pixel image blocks of the modified image, the respective multi-pixel image blocks respectively including multiple pixels of the modified image, the respective proportions representative of quantities of pixels having the first pixel value in the respective multi-pixel image blocks of the modified image;
• select first ones of the multi-pixel image blocks based on the respective proportions, the first ones of the multi-pixel image blocks to be processed to detect one or more objects;
• and
• skip processing of second ones of the multi-pixel image blocks that are not selected based on the respective proportions.