Dissertation > Excellent graduate degree dissertation topics show

Automatic segmentation and extraction of the raster city map in road

Author: WuHongLing
Tutor: WangYunQiong；FengQiaoSheng
School: Yunnan Normal University
Course: Computer Software and Theory
Keywords: Raster map Road Extraction Color Space Template matching Mathematical Morphology Skeletonization
CLC: TP391.41
Type: Master's thesis
Year: 2007
Downloads: 32
Quote: 0
Read: Download Dissertation

Abstract


In recent years, the rapid development of the geographic information system, in people's lives and the many aspects of travel, urban planning and management, traffic management, and military have been widely used. Traditional digital map manually input this way has far can not meet the demand, the map scanning input with automatic extraction of features more and more attention. The urban map information to identify the geographic information system based on road to identify and extract the important part of the city map information identified. To road extraction carried out in the the scanned raster city map, said the names of the roads, the map in the presence of a large number of various building labeling text, some text in the area, the road in some text, some text across the road and area two copies. Therefore, text, etc. The quality of the noise removal is the key extracted. In this paper, on the basis of the results of the road identification, a new road extraction method. The paper is divided into the following four parts. (1) the removal of the distinctive emblems of the template matching method based on feature points: a map image, there are a large number of icons to indicate special meaning, such as hospital signs, telecommunications flag. Hospital signs there are white areas, in line with the white road color affect road extraction. Before road extraction, template matching method to filter out affect road identification signs, road recognition accuracy rate. (2) the method of removing the text according to the LAB color difference: scan the entire image, detected black pixels, respectively, detected in a different direction. Text strokes is far less than the road or the width of the area, so in each direction simply detection fixed pixels can be. In the detection process, the step size is smaller than the threshold value, the stop when the detected road pixels, calculate the color difference detected in the direction of the pixel with the road, if the color difference is within the threshold, put these pixels into the color of the current road otherwise, sequentially detects the other direction. If all the detection direction can not be determined whether this pixel is detected on the road, and the next pixel. (3) of the road model based on the HSV space Extraction Method: in the map image on the basis of removing the noise, the image is converted from RGB color space to HSV color space, computing the H, S, V components of each pixel point According to these three components, it is determined whether the pixel belonging to road or region. Then the image closing operation with smooth boundary road primary model. (4) fusion conditions thinning algorithm and morphological thinning algorithm to extract road centerline: the conditional thinning algorithm is a classic refinement algorithm refinement speed to maintain connectivity shortcomings can not always get single refinement line of pixels wide. The template matching refinement can get a single pixel width of the image, but can not guarantee the connectivity of the image. The combination of these two methods, an effective thinning algorithm. First thinning algorithm refinement road condition, and then select a specific template image correction and removal of excess pixels get single pixels wide, connectivity refinement images.

Related Dissertations

  1. Tongue Feature Extraction and Research of Fusion Classification,TP391.41
  2. Road extraction algorithm based on region segmentation of remote sensing image,TP751
  3. Based on high-resolution remote sensing data mining houses information extraction,TP751
  4. Fundus Image Segmentation Based on SVM and Template Matching,TP391.41
  5. A Dome projection system design and implementation,TP391.41
  6. Spectrum Information hyperspectral imaging target detection technique,TP391.41
  7. Color medical image segmentation technology research,TP391.41
  8. Video sequences based image stabilization and target tracking technology,TP391.41
  9. Model based on visual characteristics and image enhancement algorithms and performance analysis,TP391.41
  10. HD image license plate location based vehicle tracking and violation detection,TP391.41
  11. Using wavelet transform and optimize embedded block truncation coding for image compression,TP391.41
  12. Infrared detection system its key technologies,TN215
  13. Haze environment image enhancement algorithm and its application,TP391.41
  14. Mobile robot vision -based target search and tracking technology,TP391.41
  15. Research and Implementation of K- means clustering of remote sensing images and watershed segmentation algorithm,TP751
  16. Swing Detection System of Straddle-type Monorail Track Beam Based on MV,TP274
  17. Extraction Algorithm based on remote sensing images of the road,TP751
  18. Development and Application of Airport Enclosure with Long-range Laser Photoelectric Alarm System,TP277
  19. The Study on the Segmentation of Grain Particles Images,TP391.41
  20. Urban Roads Extraction Algorithm for High Resolution Remote Sensing Image,TP751
  21. Image Recognition and Application of Linear Object,TP391.41

CLC: > Industrial Technology > Automation technology,computer technology > Computing technology,computer technology > Computer applications > Information processing (information processing) > Pattern Recognition and devices > Image recognition device
© 2012 www.DissertationTopic.Net  Mobile