under-overExposedImage

Very simple way to decide if an image is normal/under/overExposed using its histogram.
The implemented method is quite stupid, it divides the histogram into 10 parts to establish the exposure of the image.
Determines whether an image is underExpose, overExpose or normal.

split the 255-len hinstogram

The function splits the 256 levels of the histogram according to the nPart variable.
The surplus of the division is placed in the middle classes so as not to disturb the marginal ones that will be used to decide.
See the two examples below to understand better. ex: nPart == 5
0-->50 51-->101 102-->152~~>153 154-->204 205-->255
ex: nPart ==10
0-->24 25-->49 50-->74 75-->99 100-->124~~>130 131-->155 156-->180 181-->205 206-->230 231-->255

decision for underExpose/overExpose/normal

To make the decision i use the comparison between the sum of the two most extreme partitions and the rest of the partitions. In case the sum of the two partitions is > of the sum of remaining partitions, check that the opposite extreme is 0. This is to avoid the possibility of detecting a white object on white background as overeExpose (see the caseError example for a better understanding)

result

overExposed
newArray: [0.0, 0.0020166016, 0.003914388, 0.011096192, 0.027355958, 0.035923664, 0.05898112, 0.09416341, 0.14127849, 0.6252702]
sum (must be ~1 ) : 1.0000000116415322
over exposed
underExposed
newArray: [0.32782716, 0.24546386, 0.21890381, 0.059091795, 0.07683187, 0.04322103, 0.028448893, 0.00021158854, 0.0, 0.0]
sum (must be ~1 ) : 1.0000000050931703
under exposed
normalImage
newArray: [0.037280273, 0.09960531, 0.12129964, 0.08436361, 0.20913167, 0.2194165, 0.0678068, 0.06981364, 0.06792236, 0.023360189]
sum (must be ~1 ) : 0.9999999925494194
normal exposed image
normalImage(errorCase)
newArray: [0.0005763172, 0.0033502642, 0.0025227892, 0.0027403096, 0.0054514674, 0.004327985, 0.006090574, 0.011909807, 0.12320181, 0.83982867]
sum (must be ~1 ) : 0.9999999933643267
normal exposed image