Advertisement
pongfactory

Untitled

Nov 26th, 2015
107
0
Never
Not a member of Pastebin yet? Sign Up, it unlocks many cool features!
C 2.79 KB | None | 0 0
  1. ...... Select you own topic that related to class materials or your research. ......
  2.  
  3. The scope of your project should include the following steps:
  4.  
  5.     Identify system to be tested
  6.     Define evaluation objectives, performance metrics, parameters
  7.     Perform some measurements (if possible)
  8.     Analyze measurement data
  9.     Simulate or model the system with what-if scenario to predict performance under various conditions
  10.     Validate your simulation or model
  11.  
  12. Please submit the following information via courseville:
  13.  
  14.     Project topic
  15.     Short description
  16.     Scope of your works
  17.         what you plan to do with this topic?
  18.         how do you plan to collect measurement data?
  19.         will you use simulation or model?
  20.     Project topic
  21.  
  22. ..................................................................................................................
  23.  
  24. "Image Vectorization of Road-Dense Satellite Datasets"
  25.  
  26.     Short description
  27.  
  28. Data extraction of geo-spatial objects from satellite images is an important processing step in enabling data analysis on geo-spatial or spatio-temporal data, which commonly composes of line (road) and polygon (area) layers. Unfortunately, the traditional method is often inaccurate in extracting roads from other objects if there are many roads in the satellite image. In this paper, we present a technique for converting satellite data (raster images), that contains several roads, in the form of pixels onto the form of spatial objects that contain lines and polygons. There are three main steps in our algorithm. First, roads are extracted from other objects by employing k-means clustering. Second, the line extraction is applied only to the road areas by using morphological operation to skeletonize the image, and then, enhancing the result by using bag packing and Ramer-Douglas-Peucker algorithms. Finally, polygons can be extracted by applying land-cover classification only to non-road objects.
  29.  
  30.     Scope of your works
  31.         what you plan to do with this topic?
  32.  
  33. The experiments showed that lines (road networks) and polygons (areas) can be accurately extracted out of satellite imagery, simultaneously.
  34.  
  35.     how do you plan to collect measurement data?
  36.  
  37. Comparison between the correct image (As if it were 100 percent accuracy) and images from my program.
  38.  
  39.     will you use simulation or model?
  40.  
  41.                         Simulation
  42.  
  43.  
  44.  
  45. ..................................................................................................................
  46.  
  47. ผมยังไม่แน่ใจว่าจะทำการจำลองอย่างไรครับ ขอรายละเอียดเพิ่มเติมครับ
  48.  
  49. ..................................................................................................................
Advertisement
Add Comment
Please, Sign In to add comment
Advertisement