An algorithm for mobile vision-based localization of skewed nutrition labels that maximizes specificity - Emerging Trends in Image Processing, Computer Vision, and Pattern Recognition - page 229

The dilate-erode corner detector is used specifically because of its high sensitivity to con-
trasted text, which is why we assume that the region is bounded by these edges contains a
large amount of text. Areas of the input image which are not in focus do not produce a large
amount of corner detection results and tend not to lie within the needed projection boundar-
ies.
Two projections are computed after the corners are detected. The projections are sums of
the true pixels for each row and column. The image row projection has an entry for each row
in the image while the image column projection has an entry for each column in the image.
The purpose of these projections is to determine boundaries for the top, botom, left, and right
boundaries of the region in which most corners lie. Each value of the projection is averaged
together and a projection threshold is set to twice the average. Once a projection threshold is
selected, the first and last indexes of each projection greater than the threshold are selected as
the boundaries of that projection.
3.3 Selection of Boundary Lines
After the four corner projections have been computed, the next step is to select the Hough lines
that are closest to the boundaries selected on the basis of the four corner projections. In two
images of Figure 6 , the four light blue (white in print) lines are the lines drawn on the basis of
the four corner projection counts. The dark blue (dark gray in print) lines show the lines de-
tected by the HT. In Figure 7 (left), the botom light blue (white in print) line is initially chosen
conservatively where the row corner projections drop below a threshold. If there is evidence
that there are some corners present after the initially selected botom lines, the bottom line is
moved as far below as possible, as shown in Figure 7 (right).
FIGURE 7 Initial boundaries (left); final boundaries (right).
When the bounded area is not perfectly rectangular, which makes integration with later
analysis where a rectangular area is expected to be less straightforward. To overcome this
problem, a rectangle is placed around the selected Hough boundary lines. After the four inter-
section coordinates are computed (see Section 3.3 for details), their components are compared
and combined to find a smallest rectangle that fits around the bounded area, as shown in Fig-
ure 7 . This rectangle is the final result of the NL localization algorithm. As was stated before,
the four corners found by the algorithm can be passed to other algorithms such as row divid-
ing, word spliting, and Optical Character Recognition (OCR). Row dividing, world splitting,
and OCR are beyond the scope of this chapter. Figure 8 shows a skewed NL localize by four
algorithm.