La base documentaire de l'IFIP

La base documentaire de l'IFIP : des centaines de documents à télécharger ou bien à commander.

Résultats 1 à 20 de 97 résultats
Rechercher une documentation
Publication Annéetrier par ordre croissant
Couverture du Porc par les chiffres

Le porc par les chiffres 2019-2020

Consulter le resumé

Les chiffres clés les plus récents des filières porcines dans le monde et l’UE (production, consommation, cheptel...) et de la filière porcine en France ; les données utiles pour se repérer tout au long de l’année et à avoir toujours sous la main : un outil indispensable à tous !

  • les échanges (import/export),
  • les élevages de porcs (cheptel/régions, commerce et signes de qualité),
  • les coûts des bâtiments, le secteur de l’aliment pour porc,
  • la sélection (truies, insémination, évolutions génétiques),
  • l’abattage (entreprises, classement des carcasses, paiement au TMP),
  • le secteur de la charcuterie (entreprises et produits),
  • la consommation des viandes et la distribution des produits du porc

sous forme de tableaux, cartes, graphiques. 

Un fichier powerpoint contenant les principaux graphiques complète la brochure ; les visuels présentant chaque maillon de la filière peuvent directement servir à la préparation d’interventions techniques. Il vous sera envoyé sur simple demande : ifip@ifip.asso.fr

Edition IFIP, 39 pages, 16 X 24

25,00 €
2019

Validation of a pan-European computed tomography procedure for determining the new EU lean meat content of pigs

Consulter le resumé

Gérard Daumas (Ifip), Mathieu Monziols (Ifip) et Daiva Ribikauskiene (University of Applied Sciences of Lithuania), 65th International Congress of Meat Science and Technology (ICOMST), 4-9 août 2019, Potsdam/Berlin, Allemagne, poster

A new EU regulation on carcass classification applies from July 2018. The new reference to calibrate the pig classifi cation methods is a lean meat percentage based on total dissection of a half-carcass (LMPtd). Manual dissection can be replaced by an unbiased computed tomography (CT) procedure. 
Daumas and Monziols (2018) calibrated a simple, accurate and robust CT method in order to make it useful internationally and at least in Europe. A trial for approval of grading methods in Lithuania gave the opportunity to validate this CT procedure

PDF icon Gérard Daumas et al., 65th ICOMST, 4-9 août 2019, Potsdam/Berlin, Allemagne, poster, PDF icon Gérard Daumas et al., 65th ICOMST, 4-9 août 2019, p. 529-530
2019

Alternative à la castration : vers une détection automatique des carcasses malodorantes ?

Consulter le resumé

Patrick Chevillon (Ifip) et Pierre-Jean Escriva (Uniporc Ouest), Réussir Porc / Tech Porc (FRA), 2019, n° 270, juillet-août, p. 6-8

L’Ifip et Uniporc Ouest s’intéressent de près à une méthode d’analyse en abattoir de scatol et d’androsténone, les deux principaux marqueurs d’odeurs de verrats reconnus scientifiquement.

PDF icon Patrick Chevillon et al., Réussir Porc / Tech Porc (FRA), 2019, n° 270, juillet-août, p. 6-8
2019

Le scanner comme outil de référence pour le classement et la composition des carcasses

Consulter le resumé

Gérard Daumas, bilan 2018, éditions IFIP, avril 2019, p. 62

Le classement des carcasses participe à la rentabilité et à l’orientation de la production. Encadré par la règlementation européenne, il est basé sur la prédiction d’une teneur en viande maigre. La définition de cette teneur a changé en 2018. La teneur en muscle de la carcasse (TMC), nouvelle référence, est mesurable par dissection manuelle ou par tomographie à rayons X, si celle-ci donne des résultats comparables à la dissection. L’objectif de l’Ifip a été de développer une méthode par tomographie comparable à la dissection totale et applicable dans toute l’UE.

PDF icon Gérard Daumas, bilan 2018, éditions IFIP, avril 2019, p. 62
2019

A pan-European computed tomography procedure for measuring the new EU lean meat content of pigs

Consulter le resumé

Gérard Daumas et Mathieu Monziols, 69th Annual meeting of the european federation of animal science (EAAP), Dubrovnik, Croatie, 27-31 août 2018, poster

A new EU regulation on carcass classification applies from July 2018. The new reference to calibrate the pig classification methods is a lean meat percentage based on total dissection (LMPtd). Manual dissection can be replaced by an unbiased computed tomography (CT) procedure. If the national pig population to be sampled has the same characteristics as the population for which a CT procedure has been previously corrected, no additional dissection is required. In most of the national applications for authorisation of classification methods the population characteristics are managed via a stratification on a fat depth, mimicking the LMP variation.
The aim of this paper is to propose a pan-European CT procedure to calibrate the pig classification methods without any additional manual dissection. A sample of 29 half-carcasses was CT scanned with 3 mm slice thickness and then fully dissected according to the EU regulation. The CT muscle volume was calculated by thresholding in the Hounsfield range 0-120. It was converted into muscle weight by applying a density of 1.04. The weight was divided by the carcass weight to obtain the lean meat percentage from CT (LMPct), in the same way as done for the LMPtd. LMPtd was regressed on LMPct. Only the slope was significant and was estimated at 0.965 (s.e.=0.002). The RMSE was 0.81. The plot of residuals against fitted values showed no pattern and no heterogeneity of variances. The main source of measurement error is the thresholding of the rind. As the thickness of the rind is very thin (2-3 mm), most of the voxels including rind are mixed voxels, either with air or with fat. Their Hounsfield values are therefore less than 0 and these voxels are classified in non-muscle. Only a few rind voxels have a Hounsfield value in the range [0-120 HU] and are thus misclassified in muscle. This is taken into account by the slope value which is slightly less than 1. The LMPtd range in the sample (53-68) covers the S+E classes (>= 55) which represent more than 90% of the EU pigs. It covers too more than 80% of all the national populations, excepted Italy. This robust CT procedure can therefore be applied in 27 Member States, by using a pan-European multiplicative factor of 0.965, without any additional national dissection.

PDF icon Gérard Daumas et Mathieu Monziols, 69th EAAP, Dubrovnik, Croatie, 27-31 août 2018, poster
2018

Retrospective by sex of the new EU lean meat content of pig carcasses

Consulter le resumé

Gérard Daumas, 69th Annual meeting of the european federation of animal science (EAAP), Dubrovnik, Croatie, 27-31 août 2018, poster

A new EU regulation on carcass classification applies from July 2018. This regulation changes the definition of the reference lean meat percentage (LMP) which becomes the LMP in the carcass (from total dissection). Nevertheless, each Member State can choose when he will update its grading methods. The aim of this work is to simulate this change on the last 5 years in France for each sex.
Conversely to most of European countries sex is registered online in France during pig classification. Statistics per sex are regularly published by the regional classification organisations.
The production of entire males started in France in 2013 (13% of the males), increased quickly and is now stable (27% of the males in 2017), around 2.5 million a year. More than 95% of entire males are classified with the classification method CSB Image-Meater® (IM) approved by the EU in 2013. The present LMP prediction equation contains two fat depths (G3 and G4) and two muscle depths (M3 and M4).
A sample of 180 pigs was uniformly stratified on sex: 60 entire males, 60 females and 60 castrated males. All cuts were CT scanned allowing to calculate the LMP. Prediction equation was calculated by a general linear model including the 4 depths as well as the interactions with sex. The stepwise procedure by using BIC selected a model with 1 fat depth (G3) and 2 muscle depths (M3 and M4), G3 and M3 coefficients depending on sex. The RMSE was 2.15. Removal of M4 only decreased the RMSE of 0.01. The fat coefficient of entire males was more than twice that of females, while that of castrates was in the middle. Differences on M3 coefficient had lower impact on LMP.
The equation was applied on the annual classification averages from 2013 to 2017. In 4 years LMP increased, but differently according to sex: +0.11 for entire males, +0.25 for females and +0.34 for castrated males. In the same period the difference between entire and castrated males decreased from 4.02 to 3.79, while the difference between entire males and females decreased from 1.37 to 1.24.
When the Image-Meater method will be updated, it seems worthwhile to consider separate slopes for entire males, females and castrated males.

PDF icon Gérard Daumas, 69th EAAP, Dubrovnik, Croatie, 27-31 août 2018, abstract, PDF icon Gérard Daumas, 69th EAAP, Dubrovnik, Croatie, 27-31 août 2018, poster
2018

Computed tomography for comparing the lean meat content of different genotypes of entire male pigs

Consulter le resumé

Gérard Dumas et al., 64rd International Congress of Meat Science and Technology (ICOMST), 12–17 août 2018, Melbourne, Australie, poster

From July 2018 the EU reference for pig grading is the lean meat percentage (LMP) based on total dissection of a half-carcass. Dissection can be replaced by an unbiased computed tomography (CT) procedure. Daumas et Monziols (2018) developed a pan-European CT procedure for determining this LMP. The hypothesis is such a procedure could be applied on very lean carcasses, issued from entire males of purebred used as terminal sires in production. The aim of this paper is to test this CT procedure to assess the LMP differences between several lean genetic groups and the three halothane genotypes.

PDF icon Gérard Dumas et al., 64rd ICOMST, 12–17 août 2018, Melbourne, Australie, poster
2018

Computed tomography for comparing the lean meat content of different genotypes of entire male pigs

Consulter le resumé

Gérard Dumas et al., 64th International Congress of Meat Science and Technology (ICOMST), 12–17 août 2018, Melbourne, Australie

In the framework of pig carcass classification a new EU regulation applying from July 2018 makes compulsory the use of a new reference. It is a lean meat percentage (LMP) based on total dissection (LMPtd) of a halfcarcass. Manual dissection can be replaced by an unbiased computed tomography (CT) procedure. Daumas et al. [1] developed a pan-European CT procedure for determining this LMP. The hypothesis is such a procedure could be applied on very lean carcasses, issued from entire males of purebred used as terminal sires in production. The aim of this paper is to test this CT procedure to assess the LMP differences between several lean genetic groups and the three halothane genotypes. 

PDF icon Gérard Dumas et al., 64th ICOMST, 12–17 août 2018, Melbourne, Australie
2018

Biais par sexe et génotype halothane dans l’estimation par visionique du pourcentage de muscle des pièces de découpe

Consulter le resumé

Poster présenté par Gérard Daumas et Mathieu Monziols, aux 50es Journées de la Recherche Porcine, 2018, 6 et 7 février 2018, Paris

L’industrie de la viande s’intéresse à la prédiction des rendements des pièces afin d’orienter la découpe aussi efficacement que possible.
Ces rendements peuvent être prédits par les variables utilisées pour la pesée et le classement des carcasses. Cependant, les facteurs ayant un effet sur la composition des carcasses peuvent avoir un effet différent sur celle des pièces et spécifique à chaque pièce. Ce travail vise à quantifier les biais par sexe et génotype halothane dans l’estimation de la composition des pièces par le système automatique de vision utilisé pour le classement des carcasses.

PDF icon Poster de Gérard Daumas et Mathieu Monziols, 50es JRP, 6 et 7 février 2018, Paris
2018

Biais par sexe et génotype halothane dans l’estimation par visionique du pourcentage de muscle des pièces de découpe

Consulter le resumé

50es Journées de la Recherche Porcine, 6 et 7 février 2018, Paris, p. 97-98, par Gérard Daumas et Mathieu Monziols, poster

Poster.

L'industrie de la viande s'intéresse à la prédiction des rendements des pièces afin d’orienter la découpe aussi efficacement que possible. Les données de la classification obligatoire des porcs dans l'Union européenne (UE) permettent cette prédiction. L'automatisation de la classification des porcs est en expansion, concernant désormais environ la moitié de l’abattage européen. La technologie de la vision a été choisie par la France en particulier.
Le sexe et le gène halothane, facteurs bien connus pour leurs effets sur la composition de la carcasse, ne sont pas inclus dans les équations actuellement autorisées par l'UE pour la classification des carcasses (Font-i-Furnols et al., 2016). Par conséquent, les équations de prédiction du pourcentage de viande maigre dans la carcasse et dans les pièces pâtissent d’écarts systématiques pour ces sous-populations. L'objectif de cet article est de quantifier ces écarts dans l'estimation par la vision automatique de la composition des pièces de porc. Dans le cas où ils seraient considérés comme importants, la prise en compte du sexe et du gène halothane pourrait alors être discutée.

Sex and halothane genotype biases in machine vision estimates of the muscle percentage of pork cuts

The pork industry often guides carcass cutting by using the carcass classification results. Nevertheless, two important factors of carcass composition sex and halothane status  are not included in the prediction equations of lean meat percentage (LMP), as authorised by the European Union (EU). Systematic deviations are thus expected. This work quantified systematic deviations by sex and halothane genotype subpopulations in machine vision estimates of pork cut composition. A sample of 250 carcasses, stratified by sex (50% castrated males and 50% females), was measured on line by the automatic vision classification method used in France. An ear sample was taken to analyse the halothane genotype (Hal): normal (NN), heterozygote (Nn) or sensitive (nn). Left sides of carcasses were cut according to the EU procedure. The four main cuts were scanned by X-ray tomography to determine their LMPs, which were then estimated via regression using six potential predictors from the automatic vision system and carcass weight. Systematic deviations were calculated for each regression by sex and Hal genotype as the mean difference between fitted and observed values. Absolute deviations by sex were lowest (0.5%) in hams and highest (1.4%) in bellies. Deviations by Hal genotype were lowest (0.3%) in shoulders and highest (0.8%) in bellies. In all models, LMPs of females and Hal Nn were underestimated, while those of castrates and Hal NN were overestimated. The deviations’ sizes by Hal genotype were about half that of those by sex. Deviations by sex could be removed by using a different intercept in the prediction equations.

PDF icon Gérard Daumas et Mathieu Monziols, 50es JRP, 6 et 7 février 2019, Paris, p. 97-98, poster
2018

Impacts de la nouvelle réglementation sur le classement des carcasses de porc

Consulter le resumé

50es Journées de la Recherche Porcine, 6 et 7 février 2018, Paris, p. 75-80, par Gérard Daumas et Mathieu Monziols

L'objectif de cet article est d'anticiper les principales conséquences du changement de la réglementation de l'UE sur la classification des carcasses de porcs, intervenu en 2017. En particulier, le changement de définition du pourcentage de viande maigre de référence prendra effet nationalement à l’occasion de la prochaine actualisation des méthodes de classement. Pour quantifier cet impact, un échantillon de comparaison entre tomographie RX et dissection a été utilisé. En 2008, 29 porcs avaient été sélectionnés dans deux abattoirs selon le sexe (50% de mâles castrés et 50% de femelles). Les demi-carcasses gauches ainsi que les quatre pièces principales de découpe avaient été analysées par tomographie à rayons X. Toutes les pièces avaient ensuite été disséquées. Les images tomographiques ont été segmentées, afin de déterminer les deux teneurs en viande maigre, l’actuelle et la future. La future référence, le pourcentage de muscle dans la carcasse, s’est avérée inférieure de 2,8% par rapport à la référence actuelle (TMP, taux de muscle dans les pièces principales), soit environ 1,7 point de pourcentage. Pour prendre en compte cet écart, la procédure tomographique française pourrait être ajustée via un facteur multiplicatif de 0,965. Cette procédure s’est révélée robuste pour les principaux facteurs influençant la classification : sexe, génotype halothane et état d’engraissement. Cela permettrait d'utiliser ce facteur d'échelle, sans nouvelles dissections, pour les essais futurs pour l'autorisation des méthodes de classement en France. Cela pourrait également être utilisé dans d'autres pays intéressés par cette procédure tomographique. Enfin, les Etats membres pouvant autoriser des critères complémentaires de détermination de la valeur commerciale, l’introduction en France d’un critère tel qu’un taux de muscle pondéré par l’importance économique des pièces pourrait renforcer la compétitivité de la filière.

Impact of the new European Union regulations on pig carcass classification

The objective of this article is to predict the main consequences of the change in European Union (EU) regulations in 2017 on the classification of pig carcasses. In particular, the definition of reference lean meat percentage will change in France the next time that classification methods are updated. To quantify this impact, computed tomography (CT) and dissection were compared. In 2008, 29 pigs were selected from two slaughterhouses according to sex (50% castrated males and 50% females). The left half-carcasses and the four main cuts were CT scanned. All the cuts were then dissected. The CT images were segmented to determine current and future lean meat percentages. The future reference (i.e., the muscle percentage of carcass) was found to be 2.8% lower than the current reference (i.e., the lean meat content in the main cuts). This change represents about 1.7 points of muscle percentage. To take this gap into account, the French CT procedure could be adjusted by a multiplicative factor of 0.965. This procedure has been found robust for the main factors influencing the classification: sex, genotype and fatness. This would make it possible to use this scaling factor, without new dissections, for future trials for the authorization of classification methods in France. The factor could also be used in other EU countries interested in this CT procedure. In addition, since EU Member States may authorize additional criteria for determining commercial value, introducing a criterion in France such as muscle content weighted by the economic value of cuts could enhance competitiveness of the French pig-production sector.

PDF icon Gérard Dumas et Mathieu Monziols, 50es JRP, 6 et 7 février 2018, Paris, p. 75-80
2018

Conséquences de l'autormatisation du classement des carcasses sur l'estimation de la composition des pièces de découpe

Consulter le resumé

Gérard Daumas et Mathieu Monziols, 49es Journées de la Recherche Porcine, Paris, 31 janvier et 1er février 2017, 37-42

L’automatisation du classement des carcasses dans les grands abattoirs français s’est accompagné notamment d’un changement de localisation des épaisseurs mesurées. Les mesures latérales sur le dos (CGM) ont été remplacées par des mesures sur la fente à la jonction rein-jambon (CSB Image-Meater, I-M). Dans un contexte très concurrentiel de commerce des pièces, il est important de quantifier l’évolution de la précision de l’estimation de la composition des pièces induite par ce changement. Un échantillon de 241 carcasses, comprenant 50 % de femelles et 50 % de mâles castrés a été mesuré par ces deux méthodes dans trois abattoirs. Après une découpe européenne normalisée, les quatre pièces majeures ont été pesées et scannées par tomographie RX. Le poids et la proportion de muscle de chacune des pièces ont été calculés après seuillage des images. Chaque proportion de muscle a été régressée, d’une part sur les deux épaisseurs CGM et, d’autre part, sur les quatre épaisseurs I-M. L’épaisseur M4 de l’I-M n’était pas significative. La hiérarchie des pièces a été conservée en termes d’écart type résiduel (ETR) croissant : jambon, épaule, longe, poitrine. L’ETR a augmenté de 15 % environ pour les membres, passant de 1,8 à 2,1 pour le jambon et de 2,0 à 2,3 pour l’épaule. Par contre, il a augmenté de l’ordre de 30 à 35 % pour le tronc, passant de 2,4 à 3,3 pour la longe et de 3,0 à 3,9 pour la poitrine. L’aide du classement des carcasses à l’orientation de la découpe devrait s’en trouver affaiblie.

ENG

Consequences of automation of carcasses classification on the estimation of the cuts composition

The automation of carcass classification in the large French slaughterhouses was accompanied by a change of location of the measured thicknesses. The lateral measurements on the back (CGM) were replaced by measurements on the splitline at the junction of the loin and ham (CSB Image-Meater, I-M). In a highly competitive context of the trade in cuts of meat, it is important to quantify the extent to which the accuracy of the composition estimate of the cuts has evolved as a result of this change. A sample of 241 carcasses, comprising 50 % of females and 50 % of castrated males, was measured by both methods in three abattoirs. After a standardized European cut the four major cuts were weighed and scanned by X-Ray tomography. Weight and muscle proportion of each cut were calculated after thresholding of images. Each proportion of muscle was regressed on the one hand on the two CGM thicknesses and, on the other hand, on the four I-M thicknesses. The thickness M4 of I-M was not significant. The hierarchy of the cuts has been preserved in terms of increasing residual standard deviation (RSD): ham, shoulder, loin, belly. The RSD increased by about 15 % for limbs, from 1.8 to 2.1 for ham and 2.0 to 2.3 for shoulder. On the other hand, it increased by about 30-35 % for the trunk, from 2.4 to 3.3 for loin and from 3.0 to 3.9 for belly. The help of carcass classification for directing cutting should turn out to be weakened.

PDF icon Gérard Daumas et Mathieu Monziols, 49es Journées de la Recherche Porcine, Paris, 31 janvier et 1er février 2017, 37-42
2017

Conséquences de l'autormatisation du classement des carcasses sur les biais par type sexuel et génotype halothane

Consulter le resumé

Gérard Daumas et al., 49es Journées de la Recherche Porcine, Paris, 31 janvier et 1er février 2017, p. 53-54, poster

Le classement des carcasses de porcs définit une norme contractuelle servant de base au paiement des éleveurs. Jusqu’à la mi-2013, les porcs étaient classés par la méthode CGM, du nom de l’appareil semi-automatique manié par un opérateur prenant des mesures dorsales. Depuis, les porcs des grands abattoirs de la zone Uniporc Ouest sont classés par la méthode Image-Meater, du nom de l’appareil automatique de vision prenant des mesures à la jonction entre longe et jambon. L’automatisation du classement des carcasses s’est ainsi accompagnée d’un changement de localisation des épaisseurs mesurées. Les écarts entre sous-populations, dont la connaissance est d’un grand intérêt dans la filière, ont donc été modifiés. L’objectif de ce travail était de quantifier ces écarts entre types sexuels et génotypes halothane, facteurs connus pour leur importance sur la composition corporelle, pour les méthodes de classement CGM et Image-Meater (I-M). Pour cela, la tomographie RX a été préférée à la dissection, compte tenu de ses nombreux avantages. Le taux de muscle des pièces (TMP) mesuré par scanner selon la procédure IFIP (TMPrx) s’est avéré très proche du TMP mesuré par dissection (Daumas et Monziols, 2016).

ENG

Consequences of automation of carcass classification on differences between sexual types and halothane genotypes

The automation of carcass classification was accompanied by a change in the location of thickness measurements. Biases by subpopulation, which the sector is very interested in knowing, have therefore been changed. The objective of this study was to quantify these biases related to sexual type and halothane genotype, known for their importance on body composition, for the classification methods CGM and CSB Image-Meater (I-M). A sample of 209 carcasses was measured by both methods in three slaughterhouses. The next day, the carcasses were cut according to the standardized European method. The four main cuts were weighed and scanned by X-Ray tomography. The weight of muscle and TMP (lean meat % in the four main cuts) were calculated after thresholding images. Each cell of the factorial design (Females-Castrates; Nn-NN) brought together about 50 carcasses. For each method, the adjusted means of sex and halothane modalities were calculated by analysis of covariance including fat and muscle thicknesses as well as interactions. No interaction was significant. The difference between adjusted means by sex increased from 1.0% with CGM to 1.8% with I-M. This difference between Hal genotypes, not significant with CGM, rose to 1.3%. Systematic deviations for sexual types could be cancelled by a differentiated intercept in the prediction equation. Deviations for halothane genotypes could be reduced by including a prediction of halothane status in the equation. These deviations could also be managed at commercial level.

PDF icon Gérard Daumas et al., 49es Journées de la Recherche Porcine, Paris, 31 janvier et 1er février 2017, p. 53-54, poster
2017

Conséquences of the future EU regulation on pig carcass classification

Consulter le resumé

Gérard Daumas et Mathieu Monziols, 63e International Congress of Meat Science and Technology, Cork, Irlande, 13-18 août 2017, p. 277-278

The aim of this paper was to evaluate the main consequences of the change of the EU regulation on pig carcass classification, in particular the change of the reference lean meat percentage (LMP). A sample of 29 pigs was selected in two abattoirs and stratified according to sex (50% castrated males and 50% females). The left half carcasses as well as the four main cuts were scanned by computed tomography (CT). All cuts were then dissected. CT images were thresholded in order to determine lean meat. The future reference (LP in the carcass) will be 2.8% lowe rthan the present one (LMP in the fourmain cuts), i.E. about 1.7 percentage points. The French CT procedure could be fitted via a multiplicative factor of 0.965. This procedure is robust to the main factors - sex, genotype, fatness - influencing the classification. This would allow to use this scaling factor without new dissections for the future trials for approval of classification methods in France. It could be used too in other countries interested in this CT procedure.

PDF icon Gérard Daumas et Mathieu Monziols, 63e ICOMST, Cork, Irlande, 13-18 août 2017
2017

Comparison of pig classification results between entire and castrated males

Consulter le resumé

Gérard Daumas, 68th Annual meeting of the European federation of animal science (EAAP), Tallinn, Estonie, 28 août-1er septembre 2017, poster

Entire male is the alternative to pig castration chosen by some groups in Europe and in France by the leader cooperative. Since 2013 the French production of entire males has been increasing. Although the advantage in carcass composition is well known precise estimates are not easily available. Statistics per sex are regularly published by the French regional classification organisations. The aim of this work is to compare the national classification results of entire males with those of castrated males.

PDF icon Gérard Daumas, 68th EAAP, Tallinn, Estonie, 28 août-1er septembre 2017, poster
2017

Vision versus reflectance technology to estimate pork cuts composition

Consulter le resumé

Gérard Daumas et Mathieu Monziols, 63e International Congress of Meat Science and Technology, Cork, Irlande, 13-18 août 2017, p. 115-116

The objective of this article is to compare the precision of two carcass classification methods to assess the lean meat percentage (LM%) of the major pork cuts. One method is automatic by vision (CSB Image-Meater®, IM) and the other is semi-automatic by reflectance (CGM). A sample of 241 carcasses, comprising 50% of females and 50% of castrated males, was measured by both methods. The left sides were cut according to the EU procedure and the four main cuts were weighed and scanned by X-Ray tomography. Each LM% was regressed on the one hand on the two CGM thicknesses and, on the other hand, on the four IM thicknesses. The thickness M4 of IM was not significant. The hierarchy of the cuts has been preserved in therms of increasing residual standard deviation (RSD): ham, shoulder, loin and belly. The RSDrose about 15% for the limbs and 30-35% for the trunk. The help of carcass classification for directing cutting should turn out to be weakened.

PDF icon Gérard Daumas et Mathieu Monziols, 63e ICOMST, Cork, Irlande, 13-18 août 2017
2017

Estimation by vision of pork cuts composition for sex and halothane genotype subpopulations

Consulter le resumé

Gérard Daumas et Mathieu Monziols, 63e International Congress of Meat Science and Technology, Cork, Irlande, 13-18 août 2017, p. 52-53

This work aims to quantify the systematic deviations for sex and halothane genotype subpopulations in the estimation by automatic vision of pork cuts composition. A sample of 208 carcasses, startified according to sex (50% castrated males and 50% females), was measured on line by the classification method CSB Image-Meater® (IM). An ear sample was analysed for halothane gene (Hal). The left sides were cut according to the EU procedure and the four main cuts were CT scanned  in order to determine the lean meat percentage (LMP) in each cut. These LMPs were regressed on six IM potential predictors and the carcass weight. For each equation were calculated the systematic deviations per level of the sex and Hal factors by mean difference between predicted and observed values. The absolute deviation per sexual types was the lowest (0.5%) in ham and the highest (1.4%) in belly. The deviations per Hal genotype ranged from 0.3% in shoulder to 0.8% in belly. In all models females and Hal heterozygotes were underestimated and vice-versa. The deviations per Hal genotype were about the half than per sex. The deviations per sex could be removed by at least a different intercept in the prediction equations.

PDF icon Gérard Daumas et Mathieu Monziols, 63e ICOMST, Cork, Irlande, 13-18 août 2017, p.
2017

Vision versus reflectance technology for deviations for sex and halothane genotype in pig carcass classification

Consulter le resumé

Gérard Daumas et al. 63e International Congress of Meat Science and Technology, Cork, Irlande, 13-18 août 2017, p. 50-51

This work aims to compare the systematic deviations for sex and halothane genotype subpopulations between pig classification methods by automatic vision and semi-automatic relfectance. A sample of 209 carcasses, stratified according to sex (50% castrated males and 50% females), was measured by CSB Image-Meater® (IM) and CGM. An car sample was analysed for halothane gene (Hal). The left sides were cut according to the EU procedure and the four main cuts were weighed and scanned by X-ray tomography in order to detemrine lean meat percentage (LMP). For each device the LMP was regressed on the fat and muscle depths included in the official equations. For each equation were calculated the systematic deviations per level of the sex and Hal factors by mean difference between predicted and observed values. The LMP overestimation for castrated males increased from 0.4% with CGM to 0.7% with IM. The LMP overestimation for non-sensitive homosygotes Hal, not significant with CGM, rose to 0.5%. Systematic deviations for sexual types could be cancelled at least by a differentiated offset in the prediction equation.

PDF icon Gérard Daumas et al. 63e ICOMST, Cork, Irlande, 13-18 août 2017, p. 50-51
2017

Conséquences de l’automatisation du classement des carcasses sur les écarts entre types sexuels et génotypes halothane

Consulter le resumé

Gérard Daumas et  al., 49es Journées de la Recherche Porcine, Paris, 31 janvier et 1er février 2017, poster

L’automatisation du classement des carcasses à la mi-2013 s’est accompagnée d’un changement de localisation des épaisseurs mesurées.
Les épaisseurs mesurées sur le dos, latéralement à la fente, ont laissé place à des épaisseurs mesurées à la jonction jambon-longe, sur la fente.
Les écarts entre sous-populations, dont la connaissance est d’un grand intérêt dans la fi lière, ont donc été modifi és. L’objectif de ce travail était de quantifier, pour les méthodes de classement CGM et CSB Image-Meater® (I-M), ces écarts entre types sexuels et génotypes halothane, facteurs connus pour leur importance sur la composition corporelle.

PDF icon Gérard Daumas et al., 49es JRP, Paris, 31 janvier et 1er février 2017, poster
2017

Vision versus reflectance technology to estimate pork cuts composition

Consulter le resumé

Gérard Dumas et Mathieu Monziols, 63e International Congress of Meat Science and Technology, Cork, Irlande, 13-18 août 2017, poster

Sorting meat cuts is an ongoing challenge. Many pork plants use weighing and carcass grading information for pre-sorting. Automation of pig carcass classification entailed a change in the location of the measured variables. Since mid-2013 in France the thicknesses measured by reflectance (CGM) on the back, laterally to the splitline, were replaced by thicknesses measured by vision (CSB Image-Meater®) at the ham-loin junction on the splitline. The objective of this work was to compare the precision of these two classification methods to predict the proportion of muscle in the main cuts.

PDF icon Gérard Dumas et Mathieu Monziols, 63e ICOMST, Cork, Irlande, 13-18 août 2017
2017

Pages