Scanner automatic dirty/clean window detection
An image from a window of a scanner is taken and analyzed to determine whether the image is relevant to dirt or debris on the window. When the image is relevant to dirt or debris, the size and amount of the dirt or debris is compared to a threshold and when the threshold is exceeded an alert is raised to have the window cleaned of the dirt or debris.
1. A method, comprising:
setting, by the scanner, a threshold value based on a detected clean window of the scanner;
obtaining, by a scanner, an image, wherein obtaining further includes obtaining the image after inferring that the scanner is presently not in use based on a lack of activity being detected for: a non-zero stable weight that is detected on a weigh scale and readings from bar codes;
down-sampling, by the scanner, the image by averaging each 2×2 block of pixels in the image into a single pixel creating a down-sampled image accounting for a resolution of the image;
sampling, by the scanner, the down-sampled image for detection of dirt or debris by sampling 3×3 arrays of pixels from the down-sampled image by maintaining a current pixel at a center pixel of each 3×3 array of pixels and obtaining eight nearest neighboring pixels that surround the center pixel, wherein sampling further includes incrementing a speckle counter when a particular center pixel value is one of: less than a lowest value and greater than a highest value based on whether the image was captured using an external light source that is external to the scanner or an internal light source that is internal to the scanner; and
determining, by the scanner, whether a window through which the image was obtained by the scanner is dirty or clean based on the sampled down-sampled image by comparing pixel values for the center pixels of each 3×3 array of pixels to each of that center pixel's eight nearest neighboring pixels, incrementing the speckle counter accordingly, and comparing a speckle counter value to the threshold value for determining whether the window of the scanner is dirty or clean.
2. The method of claim 1 , wherein sampling further includes incrementing the speckle counter when a particular center pixel value is less than the lowest value and a predetermined constant value or when the particular center pixel value is more than the highest value and the predetermined value.
3. A system, comprising:
a window;
a scanner;
a dirty/clean window detector adapted and configured to: i) execute on the scanner, ii) set a threshold value for a detected clean window of the scanner and obtain an image from the scanner when the scanner determines that there is a lack of activity on the scanner for: a non-zero stable weight that is detected on a weigh scale and readings from bar codes, and iii) determine whether the image contains dirt or debris deposited on the window by processing to: a) down-sample the image, averaging each 2×2 block of pixels in the image into a single pixel creating a down-sampled image accounting for a resolution of the image, b) sample 3×3 arrays of pixels from the down-sampled image by maintaining a current pixel at a center pixel of each 3×3 array of pixels and obtaining eight nearest neighboring pixels that surround the center pixel and comparing pixel values for the center pixels of each 3×3 array of pixels to each of that center pixel's eight nearest neighboring pixels, c) increment a speckle counter when a particular center pixel value is less than a lowest value or greater than a highest value based on whether the image was captured using an external light source that is external to the scanner or an internal light source that is internal to the scanner, and d) determine the window to be dirty when the speckle counter is greater than the threshold value set for the clean window; and
an alert manager adapted and configured to: i) execute on the scanner, ii) receive an indication from the dirty/clean window detector as to whether the window is dirty or clean, and iii) turn on an alert when the window is dirty and turn off the alert when the window is clean.
4. The system of claim 3 , wherein the dirty/clean window detector is further adapted in iii) for the dirty/clean window detector to use a clean speckle count for the clean window as the threshold value.
5. The system of claim 3 , wherein the alert manager is further adapted and configured to turn off the alert when the dirty/clean window detector indicates the window is clean.
6. The system of claim 3 , wherein the system is integrated into a checkout station, and the checkout station is one of cashier operated and customer operated, and the scanner is a bar code scanner.