Algorithm 1. DBSCAN process.

[Step 1] Initialize Ptsmin = 3, ε = 1, DBSCANct = 0 and density $d=\left(W\times H\right)/255$ where W and H denote the width and the height of the region of interest. [Step 2] Let Da be the frequency of intensity value as the number of pixels. For every k-th intensity value, perform step 3 and step 4 until reaching the highest intensity value in the region of interest. [Step 3] Among the high intensity area from histogram analysis, make clusters within radius ε with consecutive intensity values larger than d; if the cluster satisfies Ptsmin threshold for that cluster, increase DBSCANct by 1. Otherwise, remove all pixels in that candidate cluster. [Step 4] Among the low intensity area from the histogram analysis, make clusters within radius ε with consecutive intensity values no larger than d; if the cluster satisfies the Ptsmin threshold, increase DBSCANst by 1. OItherwise, remove all pixels in that candidate cluster.