Can Artificial Intelligence build Future of Poultry Industry?
Poultry farming is no more relentless, it reported 1% growth in production for the year 2018.[1] Digital technologies are rising to help farmers achieve efficiency. The only question is how soon it happens.
Domestic fowl especially eggs have been a universal acceptance irrespective of culture and religion. It is estimated that 66% of the egg production is from Asia. China is accounting for 35% of global production of eggs.[2]Poultry has a far-fetched future in a country like India where GDP is linked to Protein intake.
Drones in Poultry
Although there is concern that drones could make flock nervous and stress, yet it is a better solution for free-range or yard farms where birds are allowed to roam freely. They work as nannies for these domestic fowls.
Constant monitoring of aviculture with the help of drones is beneficial to increase yield. Drones take pictures of the bird throughout the day at designated pace and send it instantly to the systems and build the database for analysing the behaviour of the flock.
How is this data used?
Developers use the data sent by drones to train computer vision models to understand the behaviour of the domesticated birds.
Training models to detect bird flu or avian diseases
The infected birds have symptoms like nasal and saliva secretions apart from high temperature. Pictures sent from drone are annotated and fed to the machines to recognise the disease at an early stage as the infection is contagious and leads to reduced production. It also helps farmers in saving the life of chicks as clogging blocks the nose of the bird leading to death.
By annotating the nasal or oral secretion on the beak of chicken with polygon tool, machines can be trained to identify the difference between a healthy bird and influenced bird. Polygon tool helps the developer to annotate precisely the target shape and texture of the secretion.
Training models to detect nutritional deficiencies in chicks.
Like human’s domestic fowl also suffers with nutritional deficiencies and young chicks die due to immobility and suffocation. A farmer can avoid such loss by early detection. Developing machines that detect bone deformities and decreased growth of young chickens need appropriate data annotation tools.
By drawing a box over the bird with box annotating tool, machines can be trained to identify the bone deformities and reduced growth of the birds. The machine is trained to understand the difference between a healthy and defected birds.
Training models to detect Behavioural diseases like cannibalism (or aggressive pecking).
Pecking behaviour often leads to the plumage of laying hens and this can lead to health and welfare conditions of the flock. Death occurs within 10 minutes of pecking, so early detection becomes the need of the hour. Building machine which gives alerts to the farmers regarding cannibalization within the timeframe of destruction will enhance the productivity of the poultry.
By Using box annotating tool, machines can be trained to detect the behaviour of feather pecking in birds and its severity. It also enables to identify the injured birds.
References:
- Economic Times: https://economictimes.indiatimes.com/news/economy/agriculture/icra-predicts-decent-growth-for-domestic-poultry-industry/articleshow/63613459.cms
2. Times of India: https://timesofindia.indiatimes.com/business/india-business/india-fastest-growing-egg-producer-in-world/articleshow/61730376.cms
To know more about OCLAVI
Signup @ https://oclavi.com/signup
Login @ https://oclavi.com/login
Social Media
