講演情報
[20a-A308-4]Intelligent identification of pollutant types and concentrations based on image processing and optical technology.
〇(M2C)Tsung-Ta Chan1, Chih-Chung Wang1, Hung-Wei Hsu1, Chao-Ching Ho2, Feng-Sheng Kao3, Jen-You Chu3, Jia-Han Li1 (1.National Taiwan Univ., 2.National Taipei Univ. of Technology, 3.Industrial Technology Research Inst.)
キーワード：
pollutant、image processing、time of flight
In conventional wastewater treatment plants, when the concentration of foam exceeds a certain level, it is often necessary to extract and send the foam for water quality testing. This process requires significant effort and resources. To deal with this issue, we have utilized the foam generated during the concentration of wastewater and applied optical and imaging technologies for intelligent identification of foam pollutants. By incorporating automated visual perception, control systems, and Internet of Things (IoT) technologies, we aim to enhance the efficiency and accuracy of water pollution detection.
We have selected samples that are prone to foaming in wastewater, such as Bovine Serum Albumin (BSA) representing biological proteins and Sodium dodecyl benzene sulfonate (SDBS) representing surfactants, and explored the variations in foam generation patterns. In the optical aspect, time-of-flight (ToF) measurements were employed to detect the stacking height of liquid foam. In the imaging part, foam features were captured using a camera, and a series of image processing techniques including grayscale conversion, sharpening, contrast enhancement, and median filtering were applied to analyze the captured foam images. Subsequently, binary processing was performed to calculate and select the foam coverage area, enabling the identification of foam types.
Finally, we will integrate all the result to achieve rapid, accurate, and non-contact detection of wastewater treatment plants. This system will assist wastewater treatment personnel in their monitoring tasks.
We have selected samples that are prone to foaming in wastewater, such as Bovine Serum Albumin (BSA) representing biological proteins and Sodium dodecyl benzene sulfonate (SDBS) representing surfactants, and explored the variations in foam generation patterns. In the optical aspect, time-of-flight (ToF) measurements were employed to detect the stacking height of liquid foam. In the imaging part, foam features were captured using a camera, and a series of image processing techniques including grayscale conversion, sharpening, contrast enhancement, and median filtering were applied to analyze the captured foam images. Subsequently, binary processing was performed to calculate and select the foam coverage area, enabling the identification of foam types.
Finally, we will integrate all the result to achieve rapid, accurate, and non-contact detection of wastewater treatment plants. This system will assist wastewater treatment personnel in their monitoring tasks.