Detection of blood spots and dirt stains in eggs using computer vision and neural networks
Patel, V.C. | McClendon, R.W. | Goodrum, J.W.
A computer imaging system was used to obtain gray level images of grade A eggs, eggs with blood spots, and eggs with dirt stains. Histograms based on the number of pixels at each gray level intensity were generated from the images and used to train neural network models. After training, the neural networks were evaluated on an independent testing set. The neural network models for blood spot detection had an accuracy of 85.6% (91.1% on grade A eggs and 86.7% on blood spotted eggs). The neural network model for dirt stain detection had an accuracy of 80.0% (88.9% on grade A eggs and 71.1% on dirt stained eggs). These accuracy levels were sufficient to produce final graded samples which were close to USDA requirements.
Показать больше [+] Меньше [-]Ключевые слова АГРОВОК
Библиографическая информация
Эту запись предоставил National Agricultural Library