<![CDATA[Indonesia is a maritime country that has a large territorial water, one of the efforts to protect the ship's hull from corrosion is by using coating technology. In this research, the process of characterizing the results of image processing on the hull was carried out to determine the corrosion level of the hull. 4 stages of the process, including: Taking samples and images of the layers of the ship parts that need maintenance and coating. In making a prototype, the coating results are assembled using the mini PCNVIDIA Jatson Nano using a webcam, then the images obtained will be processed using Edge detection uses a cany to obtain the contours of the bilge cross section of the ship. Next, using the Neural Network as an artificial material to create images captured from the captured prototype results on the results of coating or coating on the observed parts of the ship. The results of various image captures are processed and observed for shapes, patterns, contours, corrosion and coatings that are formed. The image processing method can be used as an inspection method for coating results with readings on data and software with 2 readings on data, namely reject or accept. Gradient values are related to changes in MSE, so gradient values cannot be used as a reference to justify model performance. With the addition of iterations, the value of the neural network and neural targets produced is linear.]]>
Copyrights © 2023
