Group Projects

This project was presented to the Yarl Geeks , MQED challenge . The task was to create a methodology to provide a plant's sustainability parameter externally and survival of the plant was tested.

The incubator we created was capable of providing grow light which stimulates plant growth and appropriate for photosynthesis by emitting blue and red light , in the visible light spectrum, moist, and temperature according to the plant and monitor them as well as control them. The required level(levels you mean parameters?if so better change as 'the external requirements for the growth selected plant varieties were initially...) of plants were initially collected manually and set as default in the mobile application. The data collected was directly transmitted from the incubator to our back-end (back end what?)via internet of things (IoT). With the data collected from the incubator it was decided to use machine learning to provide the optimal growth parameters for the plant. The notable factor is that all sensors needed for this project were made by us than buying off the shelf.

The team is still working towards AI applications in the field of agriculture. eg - crop dusting drone

This project was initially a self project which turned into a group project for Sri Lankan IOT competitions(SLIOT) and emerged 2nd place in the University Category .

The product is a device which is connected to a patient / client's wrist and it will collect PPG(put the long term and acronym within brackets) and temperature data and transmit to a central hub which will be placed in hospital ward,via Bluetooth low energy. The central hub will analyse the data and calculate the heart rate and the respiration rate,which will be transmitted to the back end via IoT, where data can be monitored by a doctor using a mobile application. Additional featured of a decision tree based ,Cardio vascular disease prediction will also happen(sentence doesn't make sense).

The decision tree will use demographic data such as patient age, height, weight etc.(explain the purpose of collecting demographics here) The end goal of the project was to create a centralized medical record for all citizens in Sri Lanka.

Types of forgeries targeted

-cloning -resampling -splicing

The objectives of this project

• Review and understand the literature that is published under Digital Image Forgery Detection.
• Develop fully functional modules to detect image forgery occurred with Image Cloning, Re-Sampling, Splicing and using Smoothing and Sharpening Filters.
• Finding the relevant datasets.
• Test Accuracy with the different approaches that are taken under same Image Forgery Type.

An image of 256*256 was downsampled by a scale of 2. The ISA and other implementation details can be found in my github(put the link to github within brackets). A python front end as well as matlab front end was created to send and receive the image in UART communication.

An ECG instrumentation amplifier was developed using DRL. The CMRR calculations were initially done to identify the best gain and opamp (?) configuration.