Bibliographic citations
This is an automatically generated citacion. Modify it if you see fit
Hernandez, J., (2024). Determinación del sexo de pollos In ovo con uso de imágenes hiperespectrales en reproductoras criollas [Universidad Nacional de Trujillo]. https://hdl.handle.net/20.500.14414/22225
Hernandez, J., Determinación del sexo de pollos In ovo con uso de imágenes hiperespectrales en reproductoras criollas []. PE: Universidad Nacional de Trujillo; 2024. https://hdl.handle.net/20.500.14414/22225
@mastersthesis{renati/883811,
title = "Determinación del sexo de pollos In ovo con uso de imágenes hiperespectrales en reproductoras criollas",
author = "Hernandez Valdez, Juan Ernesto",
publisher = "Universidad Nacional de Trujillo",
year = "2024"
}
Title: Determinación del sexo de pollos In ovo con uso de imágenes hiperespectrales en reproductoras criollas
Authors(s): Hernandez Valdez, Juan Ernesto
Advisor(s): Honorio Javes, Cesar Eduardo; Siche Jara, Raúl Benito
OCDE field: https://purl.org/pe-repo/ocde/ford#4.01.00
Issue Date: 2024
Institution: Universidad Nacional de Trujillo
Abstract: El objetivo del presente estudio fue la determinación del sexo de pollos in ovo con uso de imágenes hiperespectrales en reproductoras criollas. Se utilizaron 100 huevos de reproductoras criollas, de los cuales se les tomo las imágenes los días de incubación 1,3, 5, 7,8, y 10. El análisis de los datos fue realizado en el software Matlab (MATrix LABoratory) 2020a. Se emplearon pruebas estadísticas PCA (análisis de componentes principales), demostraron que hay diferencias entre los días de incubación, teniendo en cuenta el centrado de la media, así mismo se realizó Análisis Discriminante Mínimos Cuadrados (PLS-DA). Se realizó para clasificar muestras según su sexo (hembra–macho) por cada día de evaluación, se elaboró 6 modelos uno por cada día de evaluación, allí se obtuvo parámetros de sensibilidad, especificidad, precisión y tasa de error del mejor preprocesamiento obteniéndose mejor resultado en el día uno con los valores más altos de sensibilidad especificidad y precisión 33,30%, 100%, 66,7% para la clase macho, 100%, 33%, 66,67% para la clase hembra. También se realizó para confirmar predicción del sexo de muestras de huevos PLS-R (regresión de mínimos cuadrados parciales) se analizaron dos preprocesamiento para cada modelo (MC, MC+SNV (variable normal estándar)), siendo el mayor coeficiente de determinación (Rp2) 0.879 indicando una buena correlación, y para el modelo reducido (Rc2 = 0,89). Las imágenes hiperespectrales permitieron la identificación del sexo con un 66,7 % de precisión para la clase machos y 66,67% para la clase hembras, en el primer día.
The objective of the present study was the determination of the sex of chickens in ovo with the use of hyperspectral images in Creole breeders. 100 eggs from Creole breeders were used, of which images were taken on incubation days 1, 3, 5, 7, and 10. The data analysis was carried out in the Matlab software (MATrix LABoratory) 2020a. PCA (principal component analysis) statistical tests were used, demonstrating that there are differences between the days of incubation, taking into account the centering of the mean, and Least Squares Discriminant Analysis (PLS-DA) was also performed. It was carried out to classify samples according to their sex (female-male) for each day of evaluation, 6 models were created, one for each day of evaluation, there parameters of sensitivity, specificity, precision and error rate of the best preprocessing were obtained, obtaining the best result. on day one with the highest values of sensitivity, specificity and accuracy 33,30%, 100%, 66,7% for the male class, 100%, 33%, 66,67% for the female class. To confirm the prediction of the sexing of egg samples, PLS-R (partial least squares regression) was also carried out, two preprocessing processes were analyzed for each model (MC, MC+SNV (standard normal variable)), with the highest coefficient of determination (Rp2) 0,879 indicating a good correlation, y for the reduced model (Rc2 = 0, 89). The hyperspectral images allowed the identification of sex with 66, 7% accuracy for the male class and 66,67 % for the female class on day 1.
The objective of the present study was the determination of the sex of chickens in ovo with the use of hyperspectral images in Creole breeders. 100 eggs from Creole breeders were used, of which images were taken on incubation days 1, 3, 5, 7, and 10. The data analysis was carried out in the Matlab software (MATrix LABoratory) 2020a. PCA (principal component analysis) statistical tests were used, demonstrating that there are differences between the days of incubation, taking into account the centering of the mean, and Least Squares Discriminant Analysis (PLS-DA) was also performed. It was carried out to classify samples according to their sex (female-male) for each day of evaluation, 6 models were created, one for each day of evaluation, there parameters of sensitivity, specificity, precision and error rate of the best preprocessing were obtained, obtaining the best result. on day one with the highest values of sensitivity, specificity and accuracy 33,30%, 100%, 66,7% for the male class, 100%, 33%, 66,67% for the female class. To confirm the prediction of the sexing of egg samples, PLS-R (partial least squares regression) was also carried out, two preprocessing processes were analyzed for each model (MC, MC+SNV (standard normal variable)), with the highest coefficient of determination (Rp2) 0,879 indicating a good correlation, y for the reduced model (Rc2 = 0, 89). The hyperspectral images allowed the identification of sex with 66, 7% accuracy for the male class and 66,67 % for the female class on day 1.
Link to repository: https://hdl.handle.net/20.500.14414/22225
Discipline: Maestría en Ciencias Agropecuarias, mención Producción y Sanidad Animal
Grade or title grantor: Universidad Nacional de Trujillo. Escuela de Posgrado
Grade or title: Maestro en Ciencias
Juror: Ramírez Sánchez, Julia Mercedes; Alarcón Gutiérrez, Wilman Nepalí; León Gallardo, Zara Emperatriz; Honorio Javes, César Eduardo
Register date: 7-Sep-2024
This item is licensed under a Creative Commons License