|
Student Projects Offered projects for internship students and final year graduation projects
Cheap Lab
2026-01-07
The project aims to develope 3D printed devices that are frequenty needed and used in research laboratories such as microscopes, moving stages, mechanical mounting, imaging systems, chracterization systems and interferomerers.
Advisor: Waleed Mohammed
Description
The project can host one student and it is divided into two main parts:
1- Design, realization and test of the device of interest.
2- Developing a publishable manual that can be used by others to achieve the same results.
Developing low cost active prosthetic hand for children
2025-10-29
The project aims to provide an open-source material for low-cost and practical solution for the rising need for prosthetic hand particularly for children. This need has currently raised to alarming numbers between children especially with non-ethical war crimes that are being committed in the current conflicts around the world. This human made atrocities require a mean to provide hope to many victims from scholars and scientist world wide. The project is an initiative towards building such hope.
Advisor: Waleed Mohammed
Description
The project can host 2-3 students and it is divided into three main parts: mechanical hand implementation, sensing of themuscles motion, and signal processing for ordering the motors to move the prosthetic hand forthe desired operation.
Data Fusion and AI in Indoor Smart Farming
2024-10-08
This internship focuses on the integration of data fusion techniques and artificial intelligence (AI) to enhance indoor smart farming practices. The intern will explore how to collect, analyze, and utilize diverse agricultural sensor data to monitor and optimize plant growth, resource management and operational efficiency. The objectives are: gather data from various sensors and other relevant sources; implement methods for the integration of heterogeneous data set to provide valuable insights into indoor smart farming; implement machine learning algorithms to predict plant growth, detect diseases, and automate irrigation and nutrient delivery to indoor smart farming
Advisor: Romuald Jolivot
Description
Expected Outcomes: Develop a prototype of a smart farming solution that gather sensors data, perform data fusion and use it them in a machine learning framework. Duration 3-6 months Number of students: 1 Prerequisite: data science, AI Development on Python, Pandas, Scikit-learn
Development of nutrient sensors based on light attenuation
2024-10-08
The internship focuses on the investigation of a new sensor that utilize light attenuation technique to quantify different nutrient concentrations in the soil (Nitrogen, Phosphorus, Potassium). The student will engage in the research and development of sensor prototypes aimed at accurately measuring nutrient level for smart farming. The first stage involves performing simulations using Monte Carlo or other light propagation model to estimate the potential effectiveness of these sensors in real-world applications. This will be followed by experimentation and data analysis to investigate the relation between light absorption and nutrient concentrations.
Advisor: Waleed Mohammed, Romuald Jolivot
Description
Expected outcome: Light propagation model in soil, NPK light based sensor
Duration: 3-9 months
Number of students: 1
Prerequisite: Physics/optics knowledge, programming skills,
Hardware: LED, light sensor, 3D printer,
Software: Python, Arduino
Image Processing monitoring for indoor smart farm
2024-10-08
This internship focuses on the application of image processing and artificial intelligence (AI) to enhance indoor smart farming practices. The intern will explore how to capture, analyze, and utilize data from plant images to monitor and optimize growth conditions, disease detection and resource management. The objectives are: gather image data set from various sources and from our own growing plants, perform a state of the art of the image techniques used for indoor farming, implement advanced image processing algorithms to extract meaningful information; and apply machine learning techniques to identify plant health and automate responses for irrigation and nutrient delivery.
Advisor: Romuald Jolivot
Description
Expected outcome: Image processing algorithms and AI models
Duration: 3-6 months
Number of students: 1
Prerequisite: Programming skills, deep learning knowledge, image processing
Hardware: Computer and webcam
Software: Python, Deep learning tools, OpenCV
Image processing for Smart Interactive Museum Display
2024-10-08
The overall goal of this project is to create an engaging interactive experience for users by enabling them to interact with holographic images representing real artifacts, thereby overcoming the no-touching policy applied to physical items. Currently, the holograms are displayed using a four-view acrylic pyramid and an acrylic cone. The aim of the internship is to develop algorithms that facilitate interaction with the displayed holograms. There are two main components to this interaction. First, the intern will develop algorithms using the MediaPipe library to control the rotation, zoom, and display of additional textual information related to the hologram. Second, the project will incorporate a fisheye camera positioned above the hologram display to detect user presence. Image processing algorithms will be developed to track the user’s location and rotate the hologram accordingly. Depending on the progress made during the internship, additional functionalities and enhancements may be explored to further enrich the interactive experience.
Advisor: Romuald Jolivot
Description
Expected outcome: Algorithms using MediaPipe to control hologram display. Algorithm for the detection of a visitor using fisheye camera to simulate real object in a hologram display.
Duration: 3-6 months
Number of students: 1
Prerequisite: Image processing, programming skills
Hardware: Raspberry Pi, Fisheye Camera, Webcam
Software: Python, OpenCV, AI
Development of acquisition systems for holographic-like artifacts
2024-10-08
The overall goal of this project is to create advanced acquisition systems for holographic representations of real artifacts, enabling in-depth study without physical contact. The internship will focus on developing two innovative systems for data acquisition. The first system involves designing a small robot capable of performing a 360-degree rotation around an object to gather detailed visual data. This automated approach ensures thorough documentation while adhering to the no-touching policy. The second system will leverage smartphone technology to create 360-degree photogrammetry images of the artifact. This method will allow for high-quality visual representation and analysis without the need for direct interaction with the item. Together, these systems will enhance the ability to acquire and study artifacts interactively, providing users with a rich and immersive experience while maintaining the integrity of the original objects.
Advisor: Romuald Jolivot
Description
Expected outcome: 2 acquisition systems for holographic-like artifacts
Duration: 3-6 months
Number of students: 1-2
Prerequisite: Image processing, mechanical knowledge
Hardware: Raspberry Pi, Webcam, Arduino
Software: Python, OpenCV, AI
Inverse Panorama Acquisition System
2024-09-27
The internship subject is part of the Interactive Museum Display project. The goal is to develop an object panorama acquisition system for the digitization of antique ceramics. For safety reason, the ceramics need to be carefully handled, therefore, the development of a low-cost system for the digitization of ceramics is expected. By the nature of the ceramics, object panorama (also called inverse panorama) is required. The goal is to develop a system that automatically remove the background around of the ceramics and generate a 360 degree video of the object.
Advisor: Romuald Jolivot
Description
Expected outcome: Algorithms generating an inverse panorama of the digitized ceramics.
Duration: 3-6 months
Number of students: 1
Prerequisite: Image processing, programming skills
Hardware: Laptop, cameras or smartphone Software: Python, OpenCV
AI for gas detection
2023-11-26
The goal of the project is to develop an AI platform that automatically detects selected gases. The data are obtained from a set of 6 newly developed sensors that records change in resonance spectra in time-series format. The system should perform signal processing, classification and detection on records of 7 VOC (volatile organic compounds) at different concentrations.The project is currently being extended as a matrix of 81 sensors that will be recorded using a camera, switching to image processing. New AI algorithms will be required to process the new large set of data. This work is a collaboration with a national research institute that is developing new sensors, while we are working on data processing.
Advisor: Romuald Jolivot
Description
Expected outcome: AI algorithms
Duration: 3-9 months
Number of students: 1
Prerequisite: Signal Processing background, image processing knowledge, AI, Programming skills: Matlab, Python, OpenCV, Scikit-learn
AI for wearable VOC detection on smart clothes
2023-11-26
The work package involves the development of the read out system. There a python code is to be developed on Android platform to read the images from the camera while using the built in white LED as a source of excitation. Using classification algorithm and AI, the software should be able to return the concentration with predicted accuracy that would be obtained from the training and testing stages. Finally, the produced software will be tested in both lab and field environments to estimate the efficiency, durability and accuracy of the device.
Advisor: Romuald Jolivot
Description
Expected outcome: Experimental unit to measure VOC for smart clothes
Duration: 3-9 months
Number of students: 1
Prerequisite: Programming skills, basic image processing, AI knowledge.
Software: Python, OpenCV, AI toolboxes
Multispectral scanning microscope
2023-11-15
The project aims to develop a low cost multispectral scanning microscope that records both microscopic images of the sample and spectral information. The sample is placed on x-y moving stages that are to be built locally. The stages and image acquisition are to be performed by PC interfaced with Arduino through pyformata library. The spectrometer is to be built using a grating and a webcam.
Advisor: Waleed Mohammed
Description
The work involves: 1- Building a microscope using a webcam, optical lens and 3d printing. 2- Building a low cost optical spectrometer using grating and a webcam. 3- Building an x-y moving stage using DC motors and Arduino control. 4- Setup assembly to form the desired system. 4- Image acquisition and processing to extract the microscope images as well as the optical spectrum. 5- Generate an array of images and spectrum for each point in the two dimensional scans.
Flexible waveguide sensors
2023-10-02
Use of flexible substrate as a waveguide for optical sensing of strain, heat and humidity sensing. The work focuses on utilizing available optical materials such as transparent papers and transparent PLA 3D printing for the formation of the device. The project as well utilizes the concepts of wave guiding, scattering and surface Plasmon resonance for the sensing mechanism.
Advisor: Waleed Mohammed
Description
The work involves: 1- Building basic understanding of light, wave guiding and sensing. 2- Desing and building optical components using available materials. 3- Optical characterization using spectrometry and photometers. 4- Using IoT platform for data collection and analysis.
Plastic optical fiber volatile oraganic compounds sensor
2023-10-02
Plastic optical fibers are easy to handle and are cheap in cost. The project uses plastic optical fiber as a guiding region from source to detector. Scattering region is induced on the fiber using polishing techniques to generate a loss mechanism. When volatile compounds condense on the fiber, scattering is reduced and signal increases. The change of the light intensity with the concentration of the compound determines the device sensitivity. Inctreasing the sensitivity requires optimization of the polishing process and the segment length. Several fibers can be used as well in parallel to enhance the signal to noise ratio.
Advisor: Waleed Mohammed
Description
The work involves: 1- Building basic understanding of light, wave guiding and sensing.2- Desing and building optical sensor using plastic fiber and polishing technique. 3- Using IoT platform for light launching and data collection and analysis.
Demonstration of Digital Wireless Transmissions Using Software Defined Radio with HackRF One Modules
2023-08-13
The student will demonstrate a digital wireless transmission system using software defined radio with the GNU Radio software and HackRF One hardware. Commonly used modulation schemes, e.g., quadrature amplitude modulation (QAM), will be considered. In addition to learning about the fundamentals of digital communication systems, the developed demonstration systems may later be used in laboratory experiments for teaching digital communications.
Advisor: Poompat Saengudomlert
Description
Expected outcome: A demonstration wireless transmission system that can operate using one of multiple modulation schemes, including QAM, and some experimental results with comparison to theoretical predictions; Duration: 3-6 months; Hardware: HackRF One modules; Software: GNU Radio, Python
Development of an Indoor Object Tracking System Using Visible Light Communications
2023-08-13
The student will develop and test an indoor object tracking system prototype using visible light communications (VLC) through commercial LED lamps. Its potential application is tracking a moving robot in a smart warehouse. For energy savings, dimming techniques will be implemented so that localization can be done even when different LED lamps operate at different dimming levels. The developed system will offer an alternative to GPS based tracking, which is not available for indoor environments.
Advisor: Poompat Saengudomlert
Description
Expected outcome: A prototype system with LED lamps and a mobile VLC receiver sending its location to a server that displays its location and records data; Duration: 3-6 months; Hardware: LED lamps, photodiodes, related circuit components, microcontrollers (e.g., Arduino, NodeMCU), FPGAs; Software: Verilog/VHDL, Arduino IDE, Python
1. Optical Bioimaging Device and Image Processing Algorithm for Detecting a Non-invasive Biomarker on Skin for Medical Diagnosis
2023-03-21
Skin autofluorescence (AF) is a noninvasive measure of the level of tissue accumulation of advanced glycation end products. The value of skin AF has a strong correlation with microvascular complication from hyperglycemia. The skin AF can be used to predict the microvascular complications which is one of the risk factors to develop organ damage. This project aims to develop biomedical imaging system to detect the AF to reduce the risk of microvascular complications. The system includes the development of optical bioimaging device and image processing algorithm. This proposed preventive medicine will help improve the quality of life to the patients.
Advisor: Wisarn Patchoo
Description
FPGA-Based Digital Lock-in Amplifier
2023-03-21
The detection of weak signals under complex background noise is a cutting-edge technology in scientific research, geo- logical exploration, biomedical science, military, aerospace, and other fields [1]–[3]. Relevant scholars have done a lot of fruitful research in this key field and proposed a variety of detection methods, from the detection methods of linear theory such as the conventional time domain, frequency domain, and time-frequency domain, to the detection methods of non- linear theory such as chaos and stochastic resonance. Lock-in amplifier is a mature solution with strong detection ability and high reliability in many weak signal detection schemes. This project is trying to implement Digital Lock-In Amplifier (DLIA) using FPGA and DSP builder.
Advisor: Wisarn Patchoo
Description
Development of AI-based Bakery Cashier
2022-11-09
In this project, you will design and build a semi-automatic breads scanner that can recognize different kind of breads in a bakery. The bread scanning and recognizing are done via the well established object detection procedure which is widely used in the field of image and video analysis. The system shall also be able to compute the total cost for the recognized breads. Therefore, the database and related user interface shall also be designed and developed. The object detection part of the system is meant to be implemented on the edge device such as nvidia jetson or other tinyML micro-controllers. Target industry: Bakery shops in Thailand. Keywords: Object detection, Machine learning, Edge computing.
Advisor: Dr. Pakorn Ubolkosold
Description
Expected outcome: A system prototype that can scan and recognize different type of breads available in a bakery. It can also compute the total amount of cost associated with the recognized breads.
Used tools and software: Python, Nvidia Jetson, MySQL (database)
Combination of Visible Light Communication (VLC) and Power Line Communication (PLC) allowing for data signals to be transported over in-house lighting infrastructures.
2022-10-11
Due to their capacity of switching on and off rapidly, LEDs for general lighting can also be used to transmit communication signals, a concept that is known as Visible Light Communication (VLC). General lighting uses copper wire to connect the various lamps. The concept to transport communication signals over copper wire that is simultaneous used as a power line is known as Power Line Communication (PLC). In this project, the modulation method for PLC signal will be Frequency Shift Keying (FSK), whereas VLC employs the Intensity Modulation/Direct Detection (IM/DD) format. At the start of the project, students will be given a thorough introduction about the overall system after which they start working at certain areas of the system. Whenever required, students will also be given a hands-on tutorial on software packages they may need to employ.
Advisor: Karel Sterckx
Description
Expected outcome: Students are required to simulate and create hardware (including PCB design) of certain areas of the system.Number of students: 1-2Prerequisite: Knowledge of both analogue electronics, especially RLC circuitry and transformers, digital electronics. For the latter, knowledge of FPGA and its programming is helpful though not a must.Hardware: Terasic DE0-CV Development board, Software: Intel Quartus Prime, RS DesignSpark PCB
Extending the transmission bandwidth of Light Emitting Diodes (LEDs)
2022-10-11
The bandwidth of LEDs for general lighting is about 10-20 MHz when a blue filter is place in front of a photodiode. The reason for the blue filter is that LEDs for general lighting are in fact blue lights with a yellow phosphor place in front of them. The yellow phosphor acts also as a Low Pass Filter (LPF), which causes the bandwidth of the LED to drop to about an order of magnitude. In order to extend this bandwidth with external electronic circuitry, the equivalent circuit of LEDs needs to be determined. Actually, the equivalent circuit of LEDs is known and the values of the components can be determined by means of a network analyser. However, a network analyser is expensive and the measurement method complicated. As such, a simpler procedure with less expensive measurement equipment needs to be developed. After the specifics of the LED’s equivalent circuit have been determined an electronic circuit counteracts the needs to be developed and implemented on PCB. Towards this end, hands-on training on Printed Circuit Board (PCB) design using the Electronic Design Environment (EDA) RS DesignSpark PCB will be provided. The developed circuitry will be tested with various LEDs and LED lamps.
Advisor: Karel Sterckx
Description
Expected outcome: A low-cost method to determine the values of the components in the equivalent circuit of LEDs. A PCB with circuitry that counteracts the LPF characteristic of the LED and extends the LED’s bandwidth.
|