Methods and arrangements for configuring object localizers
The present technology relates to image signal processing. One aspect of the present technology involves analyzing reference imagery (e.g., depicting a conveyor belt) gathered by a camera system to determine which parts of an image frame offer higher probabilities of—relative to other image parts—containing content items. Image blocks representing localized content items can be provided to a signal detector to decode a signal.
1. An image processing method of distinguishing content items from a conveyor belt comprising:
for an M×N pixel block within an image frame representing the conveyor belt and content items, where both M and N are positive integers,
identifying a center portion of the M×N pixel block, and dividing the center portion into a plurality of sub-blocks;
for each sub-block of the plurality of sub-blocks,
determining an average or mean pixel value representing the sub-block,
updating a histogram with the determined average or mean pixel value, the histogram representing a spatial location including the sub-block across multiple image frames,
determining a histogram peak pixel value and providing a keep-out interval around the peak pixel value;
determining whether the M×N pixel block should be provided to a detector for encoded signal detection, based on whether a majority of sub-blocks have an average or mean pixel value outside of the keep out interval; and
providing the M×N pixel block to the detector based on a result of said determining.
2. The image processing method of claim 1 in which said determining whether the M×N pixel block should be provided to a detector for encoded signal detection, is based on: i. whether at a majority of sub-blocks have an average or mean pixel value outside of the keep out interval, or ii. if the M×N block is from the first K image frames analyzed, where K comprises a positive integer.
3. The image processing method of claim 2 in which K comprises a number of image frames equal to or between 32-256 image frames.
4. The image processing method of claim 1 in which the plurality of sub-blocks comprises 16 sub-blocks, and said determining whether the M×N pixel block should be provided to a detector for encoded signal detection, is based on whether at least 10 out of the 16 sub-blocks have an average or mean pixel value outside of the keep out interval.
5. The image processing method of claim 1 in which the keep out interval comprises an interval bounded in the white and dark direction between 1-20% of the peak pixel value.
6. The image processing method of claim 1 further comprising updating the keep-out interval every 32-256 image frames.
7. The image processing method of claim 1 further comprising purging histogram values every 128-1024 image frames.
8. The image processing method of claim 1 in which the peak pixel value represents the conveyor belt.
9. The image processing method of claim 8 in which a pixel value outside of the keep out interval represents a content item.
10. A non-transitory computer readable medium comprising instructions stored therein that, when executed by one or more electronic processors, cause said one or more electronic processors to perform the following acts:
for an M×N pixel block within an image frame representing a conveyor belt and content items, where both M and N are positive integers,
identifying a center portion of the M×N pixel block, and dividing the center portion into a plurality of sub-blocks;
for each sub-block of the plurality of sub-blocks,
determining an average or mean pixel value representing the sub-block,
updating a histogram with the determined average or mean pixel value, the histogram representing a spatial location including the sub-block across multiple image frames,
determining a histogram peak pixel value and providing a keep-out interval around the peak pixel value;
determining whether the M×N pixel block should be provided to a detector for encoded signal detection, based on whether a majority of sub-blocks have an average or mean pixel value outside of the keep out interval; and
providing the M×N pixel block to the detector based on a result of said determining.
11. The non-transitory computer readable medium of claim 10 in which said determining whether the M×N pixel block should be provided to a detector for encoded signal detection, is based on: i. whether at a majority of sub-blocks have an average or mean pixel value outside of the keep out interval, or ii. if the M×N block is from the first K image frames analyzed, where K comprises a positive integer.
12. Non-transitory computer readable medium of claim 11 in which K comprises a number of image frames equal to or between 32-256 image frames.
13. The non-transitory computer readable medium of claim 10 in which the plurality of sub-blocks comprises 16 sub-blocks, and said determining whether the M×N pixel block should be provided to a detector for encoded signal detection, is based on whether at least 10 out of the 16 sub-blocks have an average or mean pixel value outside of the keep out interval.
14. The non-transitory computer readable medium of claim 10 in which the keep out interval comprises an interval bounded in the white and dark direction between 1-20% of the peak pixel value.
15. The non-transitory computer readable medium of claim 10 further comprising updating the keep-out interval every 32-256 image frames.
16. The non-transitory computer readable medium of claim 10 further comprising instructions that when executed by said one or more electronic processors cause said one or more electronic processors to perform the following act: purging histogram values every 128-1024 image frames.
17. The non-transitory computer readable medium method of claim 10 in which the peak pixel value represents the conveyor belt.
18. The non-transitory computer readable medium of claim 17 in which a pixel value outside of the keep out interval represents a content item.
19. An apparatus comprising:
a camera system to capture a plurality of image frames depicting a conveyor belt and content items conveyed by the conveyor belt:
for an M×N pixel block an image frame captured by said camera system, where both M and N are positive integers,
means for identifying a center portion of the M×N pixel block and dividing the center portion into a plurality of sub-blocks;
for each sub-block of the plurality of sub-blocks,
means for determining an average or mean pixel value representing the sub-block,
means for updating a histogram with the determined average or mean pixel value, the histogram representing a spatial location including the sub-block across multiple image frames,
mean for determining a histogram peak pixel value and providing a keep-out interval around the peak pixel value;
means for determining whether the M×N pixel block should be provided to a detector for encoded signal detection, based on whether a majority of sub-blocks have an average or mean pixel value outside of the keep out interval; and
means for communicating the M×N pixel block to the detector based on a result of said determining.
20. The apparatus of claim 10 in which said means for determining whether the M×N pixel block should be provided to a detector for encoded signal detection, is based on: i. whether at a majority of sub-blocks have an average or mean pixel value outside of the keep out interval, or ii. if the M×N block is from the first K image frames analyzed, where K comprises a positive integer.
21. The apparatus of claim 19 in which the peak pixel value represents the conveyor belt, and in which a pixel value outside of the keep out interval represents a content item.