Automatic caudophagy assessment system using artificial intelligence
INNDEO participates in the CAUDOPIG R&D project with the company Le Porc Gourmet belonging to Grupo Jorge in the AEI 2023 aid line. The project consists of the inspection of carcasses in the slaughterhouse, in order to detect if the animal has suffered caudophagy, and determine its causes through the analysis of productive data.
INNDEO’s participation in this project consists of adapting its INSPECTRA CARCASS CHECKER carcass inspection equipment by vision and artificial intelligence, to process the images captured of the tail, and be able to classify them with the absence or presence of tailphagy.
What is caudophagy?
Caudophagy or tail tailing represents one of the most serious problems in the swine industry. It is about cannibalism, or nibbling, that certain animals exert on their conspecifics on their tails. When a pig bites or chews another pig’s tail, it can cause wounds and injuries to the skin and underlying tissues, which can increase the risk of infection and disease, especially in an area so close to the spinal cord. These injuries can also be painful and affect the pig’s behavior and quality of life.
This behavior affects the well-being of the animal and represents a very significant loss of productivity in pig production and, in addition, in many cases it causes infections or even the death of the animals. Furthermore, tail removal is not approved by EU legislation.
To record tail injuries, the work of Teixeira and Boyle (2014) and Carroll and Boyle (2018) are used as a basis. First, the queue integrity classification is performed:
- [0] Entire queue
- [1] Healed or minor injuries
- [2] Evidence of chewing or puncture wounds, but no evidence of swelling
- [3] Evidence of chewing or puncture wounds with swelling and signs of possible infection
- [4] Partial tail loss
- [5] Total loss of tail
CARCASS CHECKER: AI Online Pig Carcass Inspection Equipment
CARCASS CHECKER is a new inspection equipment developed by INSPECTRA that allows pig carcasses to be inspected in the slaughterhouse using vision and artificial intelligence.
The equipment can have up to 8 cameras that capture images of the most representative areas of the carcass at different times and angles, being able to detect quality failures visible to inspectors at greater speed and with greater precision.
The equipment is designed to work in a cool area and meets the highest standards of hygienic design and resistance to cleaning operations.
Its advanced technology allows for online images of the production process to be captured and processed using artificial intelligence at a rate of more than 900 carcass/hour.
This automation of the real-time classification process is achieved through the training of a neural network from thousands of images. The images in the training bank are classified and labeled indicating the reference with which we want the system to identify the product. The more images the system receives, the better trained the neural network will be, until the system “learns” to identify the different quality defects on its own with enough examples of each defect category.
The equipment is currently used to detect fecal contamination in the carcass and to classify the sex of the animal based on the cut in the inginal area, being able to show new quality defects.
How to detect caudophagy
To detect caudophagy, one of the cameras is used, in which the animal’s tail is visualized in more detail.
It is necessary to teach the neural network by labeling images of the different possible cases and once the system is trained, the neural network will automatically classify the caudophagy of the tail, giving the result whether it is visible and, if so, the degree of tailphagy detected.
The causes of caudophagy
There are many possible causes why caudophagy can occur, from lack of food, lack of ventilation, inadequate temperature, adverse environmental conditions, the pig’s search for blood, etc.
Caudophagy can be a sign of stress, boredom or lack of environmental enrichment in the production system in which the pigs are found. Chronic stress can have negative effects on the health and well-being of animals, increasing the risk of disease and affecting productive performance.
In some cases, caudophagy may be an acquired behavior due to lack of adequate space, absence of environmental enrichment objects, inadequate diet, or the presence of stressors in the production system.
Therefore, it is important to address the underlying factors that may be causing tailphagy and provide a more suitable and enriching environment for pigs.
In the course of this project, the degree of caudophagy of the animal classified by the INPSECTRA team will be associated with different data collected in the slaughterhouse, in production on the farm, together with genetic and maternal data, in order to determine the causes of the same and try to reduce its appearance.
1. Contar el número de unidades. La tecnología de inteligencia artificial de última generación que incorporan los equipos INSPECTRA, permite el conteo de las unidades envasadas, de modo que se pueden rechazar aquellos envases que no cuenten con el número de unidades especificados en el producto