Bullets defect detection based on digital image processing using line detection and fuzzy sets

This research discussed about the detection of bullet defect that aims to help the process of quality control in bullets production. This process is examined manually or without tools allowed. It will be very difficult since the capability of human eyes have limitation of sight and vision. In solving the problem, the author uses line detection and fuzzy set as the method based on digital image processing. Line detection here to detect lines of the image and fuzzy set to analyzing its similarities. The object were analyze its similarities to the image of reference. The part of bullet that examined is only on the shell chasing which made by brass. Shell chasing laid in a place with best level of light to produce a good image. The image taken directly using clear and high resolution camera. There are some bullets caliber with the range of 5,56 until 12,7 mm. However, the examination here only on the size of 5,56 mm. The process of detection had been done by using implementation program with accuracy until 70,58% using 4^th bullet reference. It successfully applied to 32 real image data of bullet, 17 images of defect bullets and 15 image of bullets with no defect.