Leakage in water distribution network (WDN) systems is a worldwide issue with serious impacts such as loss of revenue, health, and environmental concerns. This work focuses on a new analysis technique based on analysis of multiple path signal correlation in localizing single leak in a WDN. The method solves multi-directional waves issue in a WDN and shorten the analysis time of the conventional method using time-correlation. A local WDN was explored and installed with a wireless sensor network of 6 hydrophones for field experiment. Leak signals collected from each sensor were processed and analyzed in collective through the multiple paths leak localization (MPLL) system. The estimation of leak location by the MPLL method was excellent with a low percentage of error. The potential of this method as a rapid and practical leak localization system was highlighted.
{"title":"Multiple Paths Leak Localization System in Water Distribution Network","authors":"Wen Hao Png, C. Pua, Mau-Luen Tham","doi":"10.1145/3441233.3441243","DOIUrl":"https://doi.org/10.1145/3441233.3441243","url":null,"abstract":"Leakage in water distribution network (WDN) systems is a worldwide issue with serious impacts such as loss of revenue, health, and environmental concerns. This work focuses on a new analysis technique based on analysis of multiple path signal correlation in localizing single leak in a WDN. The method solves multi-directional waves issue in a WDN and shorten the analysis time of the conventional method using time-correlation. A local WDN was explored and installed with a wireless sensor network of 6 hydrophones for field experiment. Leak signals collected from each sensor were processed and analyzed in collective through the multiple paths leak localization (MPLL) system. The estimation of leak location by the MPLL method was excellent with a low percentage of error. The potential of this method as a rapid and practical leak localization system was highlighted.","PeriodicalId":326529,"journal":{"name":"International Conference on Sensors, Signal and Image Processing","volume":null,"pages":null},"PeriodicalIF":0.0,"publicationDate":"1900-01-01","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"133236107","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":0,"RegionCategory":"","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}
A. Mraz, Y. Sekimoto, Takehiro Kashiyama, Hiroya Maeda
Many municipalities and local road authorities seek to implement automated evaluation of road damage. However, they often lack technology, know-how, and funds to afford state-of-the-art data collection equipment for collection and analysis of road deficiencies. The paper describes the development of a localized road damage detection model using the transfer learning method and assessment of its usability for training the detection model from a local road image dataset of a limited size. Localized road damage dataset is created by capturing 3,923 Czech and Slovak road images containing 5,072 instances of detected road damage using a smartphone installed on the vehicle's windshield. Then, a supervised neural network was trained using the road damage dataset labeled by experts. A pre-trained MobileNet model developed by the University of Tokyo and transfer learning method were employed to accelerate the training process and to improve the model's performance when a relatively small, localized dataset is used. Finally, the performance of the developed road damage detection model was analyzed. The results show that it is possible to capture road damage into preset classes with accuracy based on the F1-score ranging between 45% and 98%. Further improvement in the detection rate can be achieved by increasing the training dataset size. The developed road damage detection model is publicly available on https://github.com/amraz39/RoadDamage DetectorCZ and it shows the high potential of employing deep neural networks in the detection of road damage by local road agencies.
{"title":"Development of the Localized Road Damage Detection Model Using Deep Neural Network","authors":"A. Mraz, Y. Sekimoto, Takehiro Kashiyama, Hiroya Maeda","doi":"10.1145/3441233.3441235","DOIUrl":"https://doi.org/10.1145/3441233.3441235","url":null,"abstract":"Many municipalities and local road authorities seek to implement automated evaluation of road damage. However, they often lack technology, know-how, and funds to afford state-of-the-art data collection equipment for collection and analysis of road deficiencies. The paper describes the development of a localized road damage detection model using the transfer learning method and assessment of its usability for training the detection model from a local road image dataset of a limited size. Localized road damage dataset is created by capturing 3,923 Czech and Slovak road images containing 5,072 instances of detected road damage using a smartphone installed on the vehicle's windshield. Then, a supervised neural network was trained using the road damage dataset labeled by experts. A pre-trained MobileNet model developed by the University of Tokyo and transfer learning method were employed to accelerate the training process and to improve the model's performance when a relatively small, localized dataset is used. Finally, the performance of the developed road damage detection model was analyzed. The results show that it is possible to capture road damage into preset classes with accuracy based on the F1-score ranging between 45% and 98%. Further improvement in the detection rate can be achieved by increasing the training dataset size. The developed road damage detection model is publicly available on https://github.com/amraz39/RoadDamage DetectorCZ and it shows the high potential of employing deep neural networks in the detection of road damage by local road agencies.","PeriodicalId":326529,"journal":{"name":"International Conference on Sensors, Signal and Image Processing","volume":null,"pages":null},"PeriodicalIF":0.0,"publicationDate":"1900-01-01","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"114953708","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":0,"RegionCategory":"","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}
This paper focused on the development of a system that could detect and locate falls, and identify the victims in a short period of time. The system used triaxial-accelerometer in detecting a fall, signal strengths from access points in locating the position, and media access control (MAC) addresses in identifying the name of the victim. Wemos D1 was used as the microcontroller in measuring and averaging signal strengths, computing the resultant acceleration coming from the MPU6050 triaxial-accelerometer and sending the values together with the MAC address to the database. The software developed accesses the database, computes for the location, and displays the outputs to the user while sounding an alarm. To test its functionality, different categories of testing were conducted. The fall function was tested and produced a recall of 100% and a precision of 97.5%. The response time was measured by how much time it took from the event of the fall to the software displaying the location and sounding the alarm. The computed average response time was 1.1128 seconds and was considered low and fast enough. The displaying of the location was tested while considering the size of the area of the testing. The area considered had the size of 10 m by 6 m and the test produced accuracies of 82.8% and 90% on x and y axes respectively. This means that the margin of error for the x-axis was 1.72 m and 0.6 m on the y-axis. In the end, the fall detection system was able to perform its function and provide reliable output that could help elderly institution, as well as elderly people, to lessen the risks and consequences of a fall.
{"title":"Fall Detection, Location and Identification for Elderly Institution","authors":"Kristine Joyce P. Ortiz, A. I. Martin","doi":"10.1145/3441233.3441245","DOIUrl":"https://doi.org/10.1145/3441233.3441245","url":null,"abstract":"This paper focused on the development of a system that could detect and locate falls, and identify the victims in a short period of time. The system used triaxial-accelerometer in detecting a fall, signal strengths from access points in locating the position, and media access control (MAC) addresses in identifying the name of the victim. Wemos D1 was used as the microcontroller in measuring and averaging signal strengths, computing the resultant acceleration coming from the MPU6050 triaxial-accelerometer and sending the values together with the MAC address to the database. The software developed accesses the database, computes for the location, and displays the outputs to the user while sounding an alarm. To test its functionality, different categories of testing were conducted. The fall function was tested and produced a recall of 100% and a precision of 97.5%. The response time was measured by how much time it took from the event of the fall to the software displaying the location and sounding the alarm. The computed average response time was 1.1128 seconds and was considered low and fast enough. The displaying of the location was tested while considering the size of the area of the testing. The area considered had the size of 10 m by 6 m and the test produced accuracies of 82.8% and 90% on x and y axes respectively. This means that the margin of error for the x-axis was 1.72 m and 0.6 m on the y-axis. In the end, the fall detection system was able to perform its function and provide reliable output that could help elderly institution, as well as elderly people, to lessen the risks and consequences of a fall.","PeriodicalId":326529,"journal":{"name":"International Conference on Sensors, Signal and Image Processing","volume":null,"pages":null},"PeriodicalIF":0.0,"publicationDate":"1900-01-01","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"127599400","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":0,"RegionCategory":"","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}