Pollen Hunt

The most abundant pollen types in Islamabad and Rawalpindi are from 08 plant species (i.e.., Paper Mulberry, Acacia, Eucalyptus, Pines, Grasses, Cannabis, Dandelion and Alternaria). Out of these plants, Paper Mulberry shares about 97% of the total pollen and its concentration touches the extreme lim

Project Title

Pollen Hunt

Project Area of Specialization

Artificial Intelligence

Project Summary

The most abundant pollen types in Islamabad and Rawalpindi are from 08 plant species (i.e.., Paper Mulberry, Acacia, Eucalyptus, Pines, Grasses, Cannabis, Dandelion and Alternaria). Out of these plants, Paper Mulberry shares about 97% of the total pollen and its concentration touches the extreme limits of about 40,000 per cubic meter of air at the peak of the blossom season. People suffering from Asthma and other respiratory diseases experience serious difficulties due to sharp increase in pollen concentrations. According to Pakistan Meteorological Department, the total pollen count in Islamabad for the year 2018 is 43330 per meter cube approximately because of the population of Paper Mulberry Trees. Our idea is to provide a density map of already classified pollen trees to help relevant government agencies, who may use the information to employ some of the preventive actions in the areas with high concentration of pollens.

Project Objectives

Automatic detection and classification of pollen trees/plants through aerial Imagery using Deep Neural Network. The primary concern of the project is to identify "pollen trees" and generate a report.

Our main approach is to obtain the imagery using mobile phone cameras and maintain it as a data-set. This data-set is processed by using competent and effective image processing and neural network techniques. The focus is to extract various features from the data and use them to train models. Later, it may be deployed to detect and classify the different species of trees with primary objective to identify pollen trees/plants. If the classified tree belongs to pollen classifications, geotag the location and create a pollen map. A comprehensive mobile application along with the web portal will be provided to disseminate the findings and help the relevant agencies to solve this very issue.

Project Implementation Method

1. Get data from metrological department about different types of pollen allergens in the air , and provide warning to user if the pollen level crosses a threshold (This data can be seen from Pakistan Met Department website: http://www.pmd.gov.pk/rnd/rndweb/rnd_new/ )
2. Develop an app and accompanied by a web portal with the features as listed below:
   1. Ability to take a snap of the plant and identify whether it is paper mulberry or cannabis (based on third-party open-source tools or relevant research.)
   2. Estimate the density of the plant in the region and geo tag the location of the plants, and possibly upload all the info on a central portal where others can see updated pollen stats. In time, an entire community will be able to spot the pollen concentration in specific areas and actual pictures of these allergy causing plants.
   3. With this information available we can add features to enable community actions e.g. all people in F-8 sector  of Islamabad, come at an event at specific place where the density of the pollen causing plants is too much and on a self-help basis geotag the location if pollen plants are found. In addition to helping the government agencies, direct information to the residents will also be available to facilitate the community in helping themselves in handling the conditions better.

Provide this information to government institutions like Ministry of Health and environment, CDA, etc. With time we gather statistics e.g. density of certain allergy causing plant in a specific area, pollen count at certain times of the day.

Benefits of the Project

The recent statistics issued by Pakistan Institute of Medical Sciences (PIMS) estimates that more than 130,000 patients with pollen allergy symptoms registered this spring at Islamabad’s various health facilities. Therefore, classification of pollen trees/plants is the most critical task to be considered.

Our proposed solution will be state-of-the-art to help in solving or reducing the effect of the issue of pollen allergy. It will be done by the classification of pollen trees to attenuate their density and control their growth through critical information about the plants, density of pollens at a time and dissemination of the relevant information to citizens; collected by our pollen hunt mobile application and made available through a web portal.

Technical Details of Final Deliverable

Technologies: Image Processing, Mobile App Development

FInal Deliverable: Mobile Application/Web Portal

Final Deliverable of the Project

Software System

Type of Industry

Health , Telecommunication

Technologies

Artificial Intelligence(AI), Others

Sustainable Development Goals

Good Health and Well-Being for People, Industry, Innovation and Infrastructure

Required Resources

If you need this project, please contact me on contact@adikhanofficial.com