<![CDATA[In this day and age, industrial progress is increasing rapidly, thus increasing the amount of industrial waste and household waste produced by humans. The accumulation of waste is caused by the long processing time of waste. Therefore, the design of a system for classifying types of waste between organic and inorganic can help separate types of waste automatically. In the design of this system, the YOLOV3 method is used. This study has several classes, namely apples, oranges, bananas, and vegetables as the type of organic waste and the class of plastic bottles, plastic cups, and cans as the type of inorganic waste. Before the detection of waste, training was carried out on the waste dataset, which amounted to 7000 images with each class having more than 1000 images. The image that has been trained will be calculated for accuracy. The calculation in the test resulted in an accuracy of 93.7% for the type of inorganic waste. While the accuracy generated on the type of organic waste is 92%. After that, the computation time on the system is calculated and the average computation time for inorganic waste is 12.5433 seconds and the average computation time for organic waste is 15.1685 seconds. The last test that was carried out was the accuracy of waste accuracy which had the smallest result of 80% and the largest 100%.]]>
Copyrights © 2022
