DETECTION OF PINHOLES IN ALMONDS THROUGH X–RAY IMAGING
暂无分享,去创建一个
Pinhole insect damage in natural almonds (i.e., in nuts with the brown skin intact) is very difficult to detect on
line. For quality reasons, methods to detect and remove such damaged nuts are of great importance. In this study, we explored
the possibility of using X–ray imaging to detect pinhole damage in almonds by insects.
X–ray film and x–ray line–scanned images of 522 pinhole–damaged almonds were obtained. Of these film images, 505
were successfully digitized to 8 bits by use of a film scanner with a 0.173–mm detector pitch, and 499 of the 505 were collected
by a line–scan unit with 0.5–mm pitch photodiodes, again at 8 bits. The pinhole–damaged region appeared slightly darker
than the non–damaged region in X–ray negative images. A machine–recognition algorithm was developed to detect these
darker regions. The algorithm used first–order (pixel intensity) and second–order (intensity change) information to identify
the damaged region. To reduce the number of false positive results due to germ regions in high–resolution images, germ
detection and removal routines were also included. With scanned film images, the algorithm showed approximately an 81%
correct recognition ratio with only 1% false positives. With line–scanned images, 65% of the pinholes were correctly
recognized with less than 12% false positives. The algorithm was very fast and efficient, requiring only minimal computation
time. The computation rate, if implemented on line, was estimated to be 66 nuts/s, while the x–ray line scanner could achieve
a scan rate of 24 nuts/s.