Vegetation Radial Clearance
Tyagi, Sachin
Vegetation Radial Clearance by Sachin Tyagi - IIT Jodhpur Department of Computer Science and Technology 2023 - vii,5p. HB
Vegetation Radial Clearance for the electricity network is the process of finding out the distance of nearby vegetation with the help of satellite images. Vegetation distance information is required to avoid outages in the electric network by trimming the vegetation on time before it grows too close. The complete process to find the radial distance involves multiple steps, from identifying the vegetation in the satellite image, then mapping the height to the placed tree, and finding the radial distance from the span (a section of the network). The height mapping and finding distance are compute-intensive tasks, as every pixel in the image is required to map to height, and then the tree polygon distance needs to be calculated based on each pixel in that polygon.
We are focused on the two problems of mapping height to each pixel of a tree polygon and finding the distance from the nearby span. As these two problems are compute-intensive, we are working with Spark to get the performance benefits of distributed computing. We will also improve the performance by moving the implementation to the GPU with Spark.
Department of Computer Science and Technology
Computer Graphics
MTech Theses
004.678 / T977V
Vegetation Radial Clearance by Sachin Tyagi - IIT Jodhpur Department of Computer Science and Technology 2023 - vii,5p. HB
Vegetation Radial Clearance for the electricity network is the process of finding out the distance of nearby vegetation with the help of satellite images. Vegetation distance information is required to avoid outages in the electric network by trimming the vegetation on time before it grows too close. The complete process to find the radial distance involves multiple steps, from identifying the vegetation in the satellite image, then mapping the height to the placed tree, and finding the radial distance from the span (a section of the network). The height mapping and finding distance are compute-intensive tasks, as every pixel in the image is required to map to height, and then the tree polygon distance needs to be calculated based on each pixel in that polygon.
We are focused on the two problems of mapping height to each pixel of a tree polygon and finding the distance from the nearby span. As these two problems are compute-intensive, we are working with Spark to get the performance benefits of distributed computing. We will also improve the performance by moving the implementation to the GPU with Spark.
Department of Computer Science and Technology
Computer Graphics
MTech Theses
004.678 / T977V