Abstract
We present a pattern recognition framework for semantic segmentation of visual structures, that is, multi-class labelling at pixel level, and apply it to the task of segmenting organs in the eviscerated viscera from slaughtered poultry in RGB-D images. This is a step towards replacing the current strenuous manual inspection at poultry processing plants. Features are extracted from feature maps such as activation maps from a convolutional neural network (CNN). A random forest classifier assigns class probabilities, which are further refined by utilizing context in a conditional random field. The presented method is compatible with both 2D and 3D features, which allows us to explore the value of adding 3D and CNN-derived features. The dataset consists of 604 RGB-D images showing 151 unique sets of eviscerated viscera from four different perspectives. A mean Jaccard index of is achieved across the four classes of organs by using features derived from 2D, 3D and a CNN, compared to using only basic 2D image features.
Highlights
Poultry is both the most popular and the fastest rising source of meat in the USA [1]
The heart and lung classes are very small in terms of pixels. For this reason a small improvement in the precision of the unary classification may lead to a large jump in overall performance, especially if it leads to a small region of the heart to suddenly be found, allowing the conditional random field (CRF) to get a better starting point for optimizing and correcting neighboring regions
This uncertainty is seen to traverse through the framework, complicating the task of the CRF
Summary
Poultry is both the most popular and the fastest rising source of meat in the USA [1]. The health inspection at poultry processing plants involves visual inspection of the viscera, that is, the internal organs of the chicken, after it has been extracted from the carcass. The inspection is necessary in order to ensure that the evisceration has been successfully performed and to examine whether the chicken shows any signs of disease. Hearts and livers are sold for human consumption. It is important that these organs are extracted undamaged and show no signs of disease. Visual abnormalities such as inflammation makes organs such as liver and heart, if not a health risk, unappetizing to look at. Incomplete removal of viscera is a quality issue in chickens that are sold whole
Sign-in/Register to view highlights & summary Talk to us
Join us for a 30 min session where you can share your feedback and ask us any queries you have
Schedule a callDisclaimer: All third-party content on this website/platform is and will remain the property of their respective owners and is provided on "as is" basis without any warranties, express or implied. Use of third-party content does not indicate any affiliation, sponsorship with or endorsement by them. Any references to third-party content is to identify the corresponding services and shall be considered fair use under The CopyrightLaw.
