• Journal of the Institute of Convergence Signal Processing
• /
• v.23 no.3
• /
• pp.166-172
• /
• 2022

In this study, we photograph driving obstacle objects such as crosswalks, side spheres, manholes, braille blocks, partial ramps, temporary safety barriers, stairs, and inclined curb that hinder or cause inconvenience to the movement of the vulnerable using electric mobility aids. We develop an optimal AI model that classifies photographed objects and automatically recognizes them, and implement an algorithm that can efficiently determine obstacles in front of electric mobility aids. In order to enable object detection to be AI learning with high probability, the labeling form is labeled as a polygon form when building a dataset. It was developed using a Mask R-CNN model in Detectron2 framework that can detect objects labeled in the form of polygons. Image acquisition was conducted by dividing it into two groups: the general public and the transportation weak, and image information obtained in two areas of the test bed was secured. As for the parameter setting of the Mask R-CNN learning result, it was confirmed that the model learned with IMAGES_PER_BATCH: 2, BASE_LEARNING_RATE 0.001, MAX_ITERATION: 10,000 showed the highest performance at 68.532, so that the user can quickly and accurately recognize driving risks and obstacles.

• Journal of the Institute of Convergence Signal Processing
• /
• v.24 no.2
• /
• pp.119-125
• /
• 2023

In this study, we photograph driving obstacle objects such as crosswalks, side spheres, manholes, braille blocks, partial ramps, temporary safety barriers, stairs, and inclined curb that hinder or cause inconvenience to the movement of the vulnerable using electric mobility aids. We develop an optimal AI model that classifies photographed objects and automatically recognizes them, and implement an algorithm that can efficiently determine obstacles in front of electric mobility aids. In order to enable object detection to be AI learning with high probability, the labeling form is labeled as a polygon form when building a dataset. It was developed using a Mask R-CNN model in Detectron2 framework that can detect objects labeled in the form of polygons. Image acquisition was conducted by dividing it into two groups: the general public and the transportation weak, and image information obtained in two areas of the test bed was secured. As for the parameter setting of the Mask R-CNN learning result, it was confirmed that the model learned with IMAGES_PER_BATCH: 2, BASE_LEARNING_RATE 0.001, MAX_ITERATION: 10,000 showed the highest performance at 68.532, so that the user can quickly and accurately recognize driving risks and obstacles.

• Journal of Korean Society of Environmental Engineers
• /
• v.31 no.11
• /
• pp.1007-1018
• /
• 2009

Hydrocyclone is widely used in industry, because of its simplicity in design, high capacity, low maintenance and operational cost. The separation action of a hydrocyclone treating particulate slurry is a consequence of the swirling flow that produces a centrifugal force on the fluid and suspended particles. In spite of hydrocyclone have many advantage, the application for treatment of urban stormwater case study were rare. We conducted a laboratory scale study on treatable potential of micro particles using hydrocyclone filter (HCF) that was a combined modified hydrocyclone with perlite filter cartridge. Since it was not easy to use actual storm water in the scaled-down hydraulic model investigations, it was necessary to reproduce ranges of particles sizes with synthetic materials. The synthesized storm runoff was made with water and addition of particles; ion exchange resin, road sediment, commercial area manhole sediment, and silica gel particles. Experimental studies have been carried out about the particle separation performance of HCF-open system and HCF-closed system. The principal structural differences of these HCFs are underflow zone structure and vortex finder. HCF was made of acryl resin with 120 mm of diameter hydrocyclone and 250 mm of diameter filter chamber and overall height of 800 mm. To determine the removal efficiency for various influent concentrations of suspended solids (SS) and chemical oxygen demand (COD), tests were performed with different operational conditions. The operated maximum of surface loading rate was about 700 $m^3/m^2$/day for HCF-open system, and 1,200 $m^3/m^2$/day for HCF-closed system. It was found that particle removal efficiency for the HCF-closed system is better than the HCF-open system under same surface loading rate. Results showed that SS removal efficiency with the HCF-closed system improved by about 8~20% compared with HCF-open system. The average removal efficiency difference for HCF-closed system between measurement and CFD particle tracking simulation was about 4%.