Broiler stunned state detection based on an improved fast region-based convolutional neural network algorithm

Abstract

An improved fast region-based convolutional neural network (RCNN) algorithm is proposed to improve the accuracy and efficiency of recognizing broilers in a stunned state. The algorithm recognizes 3 stunned state conditions: insufficiently stunned, moderately stunned, and excessively stunned. Image samples of stunned broilers were collected from a slaughter line using an image acquisition platform. According to the format of PASCAL VOC (pattern analysis, statistical modeling, and computational learning visual object classes) dataset, a dataset for each broiler stunned state condition was obtained using an annotation tool to mark the chicken head and wing area in the original image. A rotation and flip data augmentation method was used to enhance the effectiveness of the datasets. Based on the principle of a residual network, a multi-layer residual module (MRM) was constructed to facilitate more detailed feature extraction. A model was then developed (entitled here Faster-RCNN+MRMnet) and used to detect broiler stunned state conditions. When applied to a reinforcing dataset containing 27,828 images of chickens in a stunned state, the identification accuracy of the model was 98.06%. This was significantly higher than both the established back propagation neural network model (90.11%) and another Faster-RCNN model (96.86%). The proposed algorithm can complete the inspection of the stunned state of more than 40,000 broilers per hour. The approach can be used for online inspection applications to increase efficiency, reduce labor and cost, and yield significant benefits for poultry processing plants.

A Deep Learning Model for Detecting Cage-Free Hens on the Litter Floor

“…To enhance real-time detection accuracy, the YOLO (you only look once) model has been developed as a one-stage CNN for object detection. With its end to end training and entire feature maps to predict each bounding box, it performed well on real-time behavior detection of broilers and breeders [ 16 , 17 ]. Ye et al (2020) used the CNN algorithm (YOLO + multilayer residual module (MRM)) to detect 180,000 white feather broilers per hour [ 18 ].…”

Aturdimiento eléctrico y calidad de carne del pollo beneficiado en un centro de faenamiento

“…voltaje de 15 v. Como consecuencia se obtuvo un 98.20% en la mejora de la calidad de la carne [11]. Se identificó problemas en la carne de pechuga de pollo a causa de los métodos de aturdimiento empleados, para lo cual se realizó una observación directa y empleando el método de aturdimiento eléctrico (con 180 v, 400 HZ, y aproximadamente 150 mA en 5.7 s), y el aturdimiento por gas, se pudo concluir que ambos métodos cambiaron los parámetros fisicoquímicos y bioquímicos de la carne; sin embargo, el aturdimiento eléctrico produjo mayor estrés en las aves que el aturdimiento por gas [12].…”

“…Según [11], en la detección del estado de aturdimiento de los pollos, concluye que el aturdimiento moderado tiene una precisión de 98.06%, sin embargo en la presenta investigación se tuvo determino que el aturdimiento moderado tiene una precisión de 98.52%, por lo que no concuerdo con el autor Según [26], en su estudio de calidad de la carne con aturdimiento eléctrico, concluye que la cantidad de hematomas disminuyen a un 0.13%, sin embargo en la presenta investigación la cantidad de hematomas en la carne disminuye a un 2.45%, por lo que no se concuerda con el autor.…”

Keypoint Detection for Injury Identification during Turkey Husbandry Using Neural Networks

“…Videos or images have been analyzed and used for studies on broilers including bodyweight [ 20 , 21 ], health status [ 22 , 23 ], behavior [ 24 ], flock movement [ 25 ], and locomotion/activity [ 3 , 26 ]. The technology has also been used in the poultry sector for research into butchering [ 27 ], carcass and meat monitoring [ 28 , 29 ], and egg quality analysis [ 30 , 31 ].…”