This functions makes histograms of gradients per cells in a figure. It is intendend to be the first step in the histograms of oriented gradients (HOG) method which can be used as input for object detection. The following steps of the method are not implemented and thus this function is here for experimental reasons. It might be rewritten, modified, updated or deprecated in the future
make_cell_histograms(x, nbins = 9, cellsize = 8)
| x | An image object. Must be a two-dimensional matrix |
|---|---|
| nbins | Number of bins to use for the histograms that are th inpit to HOG. |
| cellsize | Cell size for HOG. Algorithm will make squre cells of whre each side is of size cellsize. |
A data.frame where each row corresponds to cell. The data.frame has nbins + 2 columns whre the firs two columns indicate the cell position, followd by *nbins* columns corresponding to each of the histogram bins.