A method includes producing a filter mask based on a blur mask and a saliency mask and identifying locations of a plurality of bounding boxes of a plurality of objects of interest in a received image. The method also includes applying the filter mask to the received image and to the locations of the plurality of bounding boxes in the received image to remove at least one object of interest from consideration. The method further includes performing a comparison of a location of a bounding box of the plurality of bounding boxes of an object of interest remaining in consideration to a predetermined safe region of the received image and generating a validation result based on the comparison.
BACKGROUND (1) Occlusion detection involves detecting whether objects in an image are occluded by other objects (e.g., objects in an overlaid image). In some instances, it may be acceptable for an object in the image to be occluded, and in other instances, it may be unacceptable. Discerning between these instances may rely on human judgment, leading to errors or inconsistent determinations and reasoning.
1. A method comprising: producing a filter mask based on a blur mask and a saliency mask, wherein the blur mask is produced by measuring a difference between each pixel of a subset of pixels in a received image and a corresponding pixel of a blurred image produced by blurring the received image; identifying locations of a plurality of bounding boxes of a plurality of objects of interest in the received image; applying the filter mask to the received image and to the locations of the plurality of bounding boxes in the received image to remove at least one object of interest from consideration; performing a comparison of a location of a bounding box of the plurality of bounding boxes of an object of interest remaining in consideration to a predetermined safe region of the received image; and generating a validation result based on the comparison. ||
8. An apparatus comprising: a memory; and a hardware processor communicatively coupled to the memory, the hardware processor configured to: blur a subset of pixels of a received image to produce a blurred image; measure a difference between each pixel of the subset of pixels in the received image and a corresponding pixel of the blurred image to produce a blur mask; determine, using a deep neural network, a saliency mask of the subset of pixels; produce a filter mask based on the blur mask and the saliency mask; identify a boundary of a first object in the image; apply the filter mask to the boundary for the first object to determine that the first object should not be obscured; and generate an alert if a second object overlaps a portion of the boundary. ||
17. A method comprising: producing a filter mask based on a blur mask and a saliency mask; identifying locations of a plurality of bounding boxes of a plurality of objects of interest in a received image; applying the filter mask to the received image and to the locations of the plurality of bounding boxes in the received image to remove at least one object of interest from consideration; performing a comparison of a location of a bounding box of the plurality of bounding boxes of an object of interest remaining in consideration to a predetermined safe region of the received image; and generating a validation result based on the comparison, comprising determining an overlap in locations of pixels contained within the bounding box and a location of the predetermined safe region of the received image.