Computer Vision Detection of Salmon Muscle Gaping Using Convolutional Neural Network Features
- PDF / 8,089,971 Bytes
- 14 Pages / 595.276 x 790.866 pts Page_size
- 12 Downloads / 220 Views
Computer Vision Detection of Salmon Muscle Gaping Using Convolutional Neural Network Features Jun-Li Xu 1 & Da-Wen Sun 1
Received: 10 February 2017 / Accepted: 31 May 2017 # Springer Science+Business Media, LLC 2017
Abstract Salmon muscle gaping will lead to the irregular voids or undesirable lace-like appearance in the final product. This study was carried out to develop an automatic imaging analysis method for rapid, accurate, and non-invasive detection of gaping blemishes on salmon carcasses. Salmon fillets could be classified as wholesome or defective samples based on the number of candidate gaping regions in the preliminary step applying local adaptive thresholding. Supervised classification results were compared between using histograms of oriented gradients (HOG) and convolutional neural network (CNN) feature extractors. It was shown that CNN features outperformed HOG features with correct classification rates (CCRs) of 0.927 and 0.916 for cross validation and test data set, respectively. Relieff was then applied to select important feature attributes by reducing the 4096-dimensional to 239dimensional vector. Simplified CNN model also yielded good classification performance with CCR of 0.925 for cross validation. Therefore, CNNs were used to extract features from candidate regions and then reduced features to the 239dimensional vector. The resultant vector was fed to the simplified CNN model to make a final decision. The prediction maps for visualizing the classification result on salmon fillet were subsequently generated. The overall results confirmed that this proposed method is effective and suitable for the muscle gaping detection. Future work will be focused on applying this method in packing plants where fish fillets are progressing rapidly, and promising results will allow the * Da-Wen Sun [email protected]; http://www.ucd.ie/refrig; http://www.ucd.ie/sun 1
Food Refrigeration and Computerized Food Technology (FRCFT), School of Biosystems and Food Engineering, National University of Ireland, Agriculture and Food Science Centre, University College Dublin, Belfield, Dublin 4, Ireland
identification of critical points in the supply chain that impact upon product quality. Keywords Computer vision . Salmon . Muscle gaping . CNN . HOG . SVM
Introduction Aquatic products such as salmon are highly perishable, therefore common techniques such as cooling (McDonald et al. 2000; Sun 1997; Sun and Brosnan 1999; Sun and Wang 2000; Sun and Hu 2003; Wang and Sun 2002; Wang and Sun 2002; Wang and Sun 2004; Zheng and Sun 2004), freezing (Cheng et al. 2016; Cheng et al. 2017; Kiani et al. 2011; Ma et al. 2015; Pu et al. 2015; Xie et al. 2015; Xie et al. 2016) and drying (Cui et al. 2008; Pu and Sun 2016; Yang et al. 2017) are normally used to keep their quality. On the other hand, it is also important to develop novel methods for evaluating their quality. For Atlantic salmon fillets, muscle gaping is one of the major causes of their downgrading (Michie 2001). It is quite problematic because gaping results in lacelike
Data Loading...
