The aim of this study was to identify possible markers to distinguish differences between land animals by using the microscopic method in association with computer image analysis. For this purpose bone fragments from poultry and mammals were obtained and analysed by microscopic method. Through a digital camera and an image analysis software 85 bone lacunae images have been processed and elaborated in order to obtain for each lacuna a monochrome mask on which several measurements were performed. Data were analysed by ANOVA and LDA. Results obtained in the present study indicated that of 32 descriptors processed by image analysis software, only 12 were significantly (P<0.001) different between mammalian and poultry. However, when morphometric measurements were analysed by LDA, 86% of lacunae were correctly classified into the animal class of origin (i.e. mammalian as mammalian and poultry as poultry). By contrast 14% of lacunae were incorrectly classified. In conclusion, data here presented indicate that some of descriptors used by image analysis appears promising not only for a reliable distinction between the different origins of animal meal at the level of vertebrate classes, but also for further characterisation and identification of processed animal proteins in animal feeds.
Selection of new markers for animal by-products characterization by classical microscopy
CAMPAGNOLI A;
2007-01-01
Abstract
The aim of this study was to identify possible markers to distinguish differences between land animals by using the microscopic method in association with computer image analysis. For this purpose bone fragments from poultry and mammals were obtained and analysed by microscopic method. Through a digital camera and an image analysis software 85 bone lacunae images have been processed and elaborated in order to obtain for each lacuna a monochrome mask on which several measurements were performed. Data were analysed by ANOVA and LDA. Results obtained in the present study indicated that of 32 descriptors processed by image analysis software, only 12 were significantly (P<0.001) different between mammalian and poultry. However, when morphometric measurements were analysed by LDA, 86% of lacunae were correctly classified into the animal class of origin (i.e. mammalian as mammalian and poultry as poultry). By contrast 14% of lacunae were incorrectly classified. In conclusion, data here presented indicate that some of descriptors used by image analysis appears promising not only for a reliable distinction between the different origins of animal meal at the level of vertebrate classes, but also for further characterisation and identification of processed animal proteins in animal feeds.I documenti in IRIS sono protetti da copyright e tutti i diritti sono riservati, salvo diversa indicazione.