Project number
23027
Organization
Roche Tissue Diagnostics
Academic year
2022-2023
The BenchMark ULTRA, an automated tissue staining instrument used in the detection of diseases such as cancer, requires several fluid reagents to perform its analyses. Detecting and maintaining adequate bulk reagent levels currently requires technicians to manually log usage, a process which is burdensome, time consuming and subject to error.
The team developed a cart-mounted, contactless vision system to automatically sense reagent fluid levels. The vision system includes a high-quality camera connected to a Raspberry Pi. An LED strip is used in conjunction with induced light blockers on the custom designed cart to stabilize the environmental conditions and improve measurement accuracy. The Raspberry Pi processes the camera image using gamma correction and performs fluid level estimation using pixel intensity comparisons, resulting in detection of fluid levels within 5% of the true volume. The analysis results are displayed on the system graphical user interface and saved for later use. A rechargeable power supply allows for 30 minutes of uninterrupted use.
The team developed a cart-mounted, contactless vision system to automatically sense reagent fluid levels. The vision system includes a high-quality camera connected to a Raspberry Pi. An LED strip is used in conjunction with induced light blockers on the custom designed cart to stabilize the environmental conditions and improve measurement accuracy. The Raspberry Pi processes the camera image using gamma correction and performs fluid level estimation using pixel intensity comparisons, resulting in detection of fluid levels within 5% of the true volume. The analysis results are displayed on the system graphical user interface and saved for later use. A rechargeable power supply allows for 30 minutes of uninterrupted use.