Utilization of image processing in this almost digital era is very much needed in
various fields, such as in the fields of medicine, forensics, law, trade, education
and in everyday life, especially images that can convey information. Digital
images can convey information. Digital images can be processed or modifield into
other digital images, this process is called digital image processing. Image
processing has the aim of improving image quality so that it can be easily
interpreted by humans or computers. Image segmentation is an important part in
image analysis, because in this process the desired image/image will be analyzed
for further processes to make it easier to analyze. This study will try to use one of
the local thresholding methods, namely the niblack threshold method in the case
study of the perforated leaf monstera adansonii. The image segmentation process
starts from image acquisition, convert RGB to grayscale, local mean, standart
deviation, then looks for the thresholding value. The niblack threshold (local
thresholding) and global thresholding methods ned to be compared, so the error
value is searched using the mean square error (MSE). The result of the MSE test
using 20 sampel monstera adansonii got a success percentage of 75% the lowest
MSE value in the niblack threshold method was obtain in the 5 sampel with a
value 1572.2281, while the global thresholding MSE value was 7744,8932. The
niblack thresholding method in other words is better than the global thresholding
method in terms of error.