Our Projects
Project title: Development of sensing system based on foliar spectral reflectance to estimate nutrients in potato plants
“Global competition had led to develop farmer’s plans to promote the technological advancements into their production line. One of those techniques are the ground-based optical sensors, that are used to measure the agricultural crop’s reflectance. Commercial sensors are at present available to make a real-time application of nitrogen. While all other minerals are still following the conventional practices using foliar tissue analysis. These conventional practices have been widely applied due to their reliability in determining the nutritional status, however, they are time consuming, costly, and cannot meet the requirement for a quick diagnosis in the field. Thus, nutrient determination needs a rapid, simple, and cost-effective method for routine analysis.
We aim to estimate potato plant nutrition status by studying the relationship of chemical analysis of petioles and spectral analysis of leaves. The developed sensing system shall be deployed into an integrated machine learning system to estimate the nutrients in the field.”
” Precise application of pesticides in crop fields may positively impact the environment. This prompts the need to explore options such as variable rate application that may reduce pesticide inputs. The objective of this research is to develop a boom spraying system for site-specific application of pesticides. The plan is to ensure that the system applies sufficient and appropriate chemical quantities on to pest targets while overcoming spray timing problems associated with hardware limitations, desired travel speed, and external environmental conditions. Evaluation tests on key components such as pumps, valves and nozzles will be carried out at laboratory and field scale to assess their compatibility with machine vision systems. Reduction in pesticide application quantities is expected to bring about farm input savings. Pesticide efficacy will also be enhanced by avoiding co-mixing of chemicals as the parallel system to be developed will operate via an auxiliary tank.”
Project title: Towards applying precision agriculture on boom sprayer by integrating sensing and actuation components for real-time spot applications
” The overall objective of my research is to develop an on-the-go see and spray system that can detect the anomalies of potato plants without hampering them and spray the required amount of fluid in that specific area where the inconsistencies are. This research starts just after the detection of anomalies. An electronic control unit (ECU) connected with the CAN Bus will detect the specific location of the anomalies and spray on that specific spot the exact amount required. All the cumulative detection obtained will be used to create a prescription map. While spraying, all the cumulative sprayed spots will be used to generate a follow-up map. Both prescription and follow-up maps will use for analyzing the data and evaluating the result. “
Project title: Development of Real-Time Intelligent Machine Vision System to Detect the Symptoms of Potato Plants
” The aim of my research is to replace old human based scouting process with computer-aided scouting procedure. Due to advancement of technology, computer vision and deep learning has progressed manifolds and it is applied in vast fields such as object detection. Therefore, to achieve maintained yield, scouting can be aided with computer vision-based technology. A machine vision system consisting of an RGB camera that can replace in part or totally the conventional farming scouting practices is employed. It helps in real-time detection and classification of crop symptoms and diseases. Furthermore, symptom maps using state-of-the-art deep learning techniques are generated.”
Project title: Machine Vision-Based Potato Tuber Inspection Station for Quality Mapping of Storage Facility
“The aim of my research will be to design and implement an imaging system for the quality assessment of post harvest potato tubers using machine vision. These potato quality parameters will be supplemented by remote conveyor system orientation sensing and used to construct a quality map of potato tubers within a post-harvest storage facility.”
Project title: Developing sensing technique to predict sufficiency of principal nutrients in potato plants based on spectroscopy
“The research aims to determine the sufficiency and deficiency level of main nutrients which is immediately related to potato yield. After getting spectral data from potato leaves by using a spectrophotometer, and chemical results from laboratories, the data will be analyzed to find the relationship between spectral data and nutrients. To develop a model, Neural Network will be used in classification of nutrient sufficiency, which will generate fast and more accurate results. Research also aims at finding nutrient interactions, such as effects of excessive nutrients on each other. “