Food assessments and security after Natural Disasters.
iSpeech
The conjunction between big data and machine learning algorithms allows monitoring agricultural resources, which are of extreme importance in countries where agriculture is a primary source of livelihoods and a key driver of economic growth. Countries around the world are exposed to high-risk disasters including cyclones, earthquakes, tsunami, storm surge, volcanic eruptions, landslides, and droughts. Theses recurrent disasters cause damage and losses to agriculture, food security and the local economy. In the last years, according to the 2015 Report of the Secretary-General on the Implementation of the International Strategy for Disaster Reduction; disasters worldwide cost around USD 1.5 trillion in economic damage. The frequency and severity of natural disasters are increasing, revealing an urgent need to strengthen the resilience of food assessments and security.
We propose an approach that fulfills the necessities of rapid food security, assessment, planning, exploitation, and management of agricultural resources, by building a framework to localize and classify four type of fruits efficiently. Followed by a method to automatically identify and segment roads to detect the faster and safer way to transport them to the adjacent warehouse or security point. The novelty of this approach brings low-cost and time efficient method for inventory, mapping, harvesting, and management of agricultural resources, leading to a fast and accurate estimation of fruit quantity, improving fruit recollection, and assess the impact of disasters on food security.
To do so, we used two supervised deep CNNs, one known to perform the task of object localization, to localize and classify the type of trees. These trees and their locations can be compared with the ere state to understand better how local agriculture and food security were affected. This information can directly inform and accelerate subsequent relief efforts. Additionally, we propose a method to determine the density of each of these trees to improve productivity, based on the detection results and presented as Density Maps to quickly comprehend the condition of the agricultural site.
The second CCN model performs a semantic segmentation, to mask the streets from the aerial imagery to help identify local transportation infrastructure and in the scenario of natural disasters understand the impact onto them and inform the proper plan to distribute aid across affected areas. Ultimately, we introduce a method to optimize the harvesting process, based in specific sceneries, such as maximum time, path length, and location of the warehouse and security points.
In future work, inspired by the success of this experiment we could apply this approach to any question regarding localization, classification or transportation of resources, an example could be the assessment of damage (low mid and hight) in buildings after a natural disaster. Another application could be for monitoring the type of rooftop materials to determine the social condition of some regions of interest.
2018-08-15 14:40:27
source
Gloss