Thanks to the development of software with artificial intelligence created by a group of researchers from the Universidad Pontificia Bolivariana (UPB) small and medium oil palm in Sabana de Torres (Santander) know in minutes the current state of their crops.
Get accurate real state of their plantations data, if you consider that most accounting and phytosanitary control is done manually, it becomes one of the major difficulties faced by producers of palm invest up to three months and two million pesos to survey medium of 1,500 hectares crop.
Given this scenario the group consisting of Jhany Zulyma Miserque and Rubbermaid Laverde, and directed by the teacher Claudia Leonor Rueda faculty Electronic engineering at the Pontificia Universidad Bolivariana (UPB) in Bucaramanga, worked on the development and implementation of an air counting system palm plantations oil, which in its first phase yielded good results.
tests conducted by the research group managed account, using techniques of artificial vision and with 97% efficiency, large tracts of land dedicated to the production of this fruit.
the artificial vision detects light and interprets to classify specific ways in specific events, recognizing the palms and differentiating them from other objects to make a systematic process accounting.
“the process of information is done through algorithms Machine Learning (Machine Learning ), which takes data, sorts and interprets for an automated recognition of palms is made and a count of plantations is generated, “said Claudia Rueda, director of the project.
Other systems are very expensive
in the world other systems are applied as making use of satellites but the costs are very high. The advance took into account that some methods of manual counting yield inaccurate data thus lead to the palm growers work with outdated between one and two years old information.
These techniques are done counting approximately plant per square meter, according to the agronomist Fundepalma, Julian Rueda, “does not allow fully detected problems such as pests and diseases, as plantations should see above to detect these shortcomings”.
Colciencias expected to be linked to the financing of the second and third phases of the project to ensure that, through the new program, identify early diseases such as spear rot (PC) that leaves Santander losses of 2.3 billion weights for the allocation of 37,000 hectares.
According Fedepalma Colombia is the largest producer in Latin America and the fourth in the world. Until September 2015 domestic production of crude palm oil reached a record high of 1.2 tons, representing a growth of 10%, compared to the end of 2014.
However, the sector faces two problems. the first, the absence of a mechanism inventory of these large plantations and Phytophthora palmivora known as bud rot
by Diana Marcela Clavijo
TIME
BUCARAMANGA
No comments:
Post a Comment