<p dir="ltr">A fluid extraction device designed to extract bands of fluid from an ultracentrifuged vial was developed. To do this, a computer vision system was developed to scan the vial and detect the bands and calculate their positions. The computer vision was performed using a TSL1401 DB linescan camera. The vial was scanned by moving the vial vertically in front of the camera using a stepper motor driven linear stage. Pixel data from the line scan camera was analysed to find the edges of each band. This was done by analyzing the first and second derivatives of smoothed pixel data. When the second derivative was 0 and the first derivative was greater than a threshold value, a band edge was detected. Needles were connected to linear actuators and were used to puncture the vials at the desired band's position. After puncturing the vial, fluid was extracted from the vials using a peristaltic pump and placed in a reservoir. Several control architectures were tested an used to control the system. The final control architecture used a LattePanda miniPC to control the system. Several iterations of the fluid extraction device were also created, with the final iteration meeting all the requirements for the device.</p>
| Identifer | oai:union.ndltd.org:purdue.edu/oai:figshare.com:article/25676385 |
| Date | 28 April 2024 |
| Creators | Aayush Mathur (18431358) |
| Source Sets | Purdue University |
| Detected Language | English |
| Type | Text, Thesis |
| Rights | CC BY 4.0 |
| Relation | https://figshare.com/articles/thesis/Development_of_an_Autonomous_Extraction_System_for_Ultra-Centrifuged_Fluids/25676385 |
Page generated in 0.0039 seconds