Dissertation > Excellent graduate degree dissertation topics show

Segmentation and Recognition of Red Cells in Urinary Sediment Microscopic Images

Author: HuangZuoXiao
Tutor: GuoSiYu
School: Hunan University
Course: Biomedical Engineering
Keywords: Urinary sediment Erythrocyte partition ID Mean shift Hough transform Ellipse detection
CLC: TP391.41
Type: Master's thesis
Year: 2009
Downloads: 75
Quote: 2
Read: Download Dissertation

Abstract


The urinary sediment microscope check the is one of the the common rather important of the the clinical Inspection project and the, and the red blood cell is detection one of the an important indicator of of. Currently, the clinical examination of the urinary formed elements mainly rely on manual microscopy. The way work intensity, the degree of subjectivity, and mainly focus on the qualitative examination of the physical components, is not conducive to clinical quantitative diagnosis. With the development of computer vision and pattern recognition technology, automated urine sediment visible components detected quantitative analysis has become possible. The process of automatic analysis of is can be divided for to for the the image preprocessing, segmentation, Extracted and Its Recognition a characteristics of in it. This thesis is mainly on the segmentation and recognition of the red blood cells to commence the study. Mean Shift algorithm in terms of image pre-processing, color images directly smoothing, image after Canny operator edge detection, the edge of the object to obtain more complete, and to some extent inhibited the effects of background noise. Red blood cell shape is generally close to the characteristics of the ellipse, this paper used parametric curve detection method - Hough transform for segmentation of red blood cells. Due to the Description that the ellipse need to use the the the 5 parameter, Therefore the Standard Hough transform to does not has a the the feasibility of in the the In the present study, while the the randomized Hough transform is even more apply. Analysis of randomized Hough transform sampling efficiency, pointed out the need to reduce the noise level in the sampling process, and improved randomized Hough transform ellipse detection algorithm based on noise reduction ideas proposed. Based on a priori knowledge of the size of red blood cells in the whole map Hough transform ring detection certain region may exist erythrocyte random ellipse Hough detection, and then turn each candidate region. To further reduce the complexity of the algorithm in a reasonable memory overhead, the Hough transform parameter space to construct five one-dimensional accumulator array. Random Hough transform in a small land within the local signal-to-noise ratio is greatly improved, so the algorithm efficiency can be greatly improved, the accuracy rate has improved markedly. Through the experimental in the the on the the urinary sediment image, verify the a the the the validity of of the algorithm. The ellipse object obtained by dividing, for a certain geometric feature extraction and recognition decision tree constructed based on ID3 algorithm erythrocytes. The resulting decision tree in the the actual urinary sediment image set confirmatory applications, and achieved good results.

Related Dissertations

  1. Tracking Cells in High Density Image Sequences Based on Mean Shift Algorithm Combined with Topological Constraint,Q25
  2. Research on Recognition of Fabric Weave Pattern Based on Space-Frequency Domain,TS101.923
  3. Research on License Plate Location Algorithm of License Plate Recognition System,TP391.41
  4. Research on Object Detection and Tracking Method in Active Vision System,TP391.41
  5. Research on the Vision-Based Driver Fatigue Real-Time Detection,TP391.41
  6. The Study of Moving Object Detection and Tracking Algorithm Based on Image Information,TP391.41
  7. Research of Image Segmentation in Web Image Search Based on GPUs,TP391.41
  8. Based on an RFID chip placement GHT Visual Positioning Technology,TN405
  9. Video image sequence moving target acquisition and tracking,TP391.41
  10. Mean-Shift -based KLT and target tracking study,TP391.41
  11. Depth map and color images based on treadmill game interaction system,TP391.41
  12. Visual underwater target detection and recognition,TP391.41
  13. Object Detection and Accurate Localization in Industrial Application,TP391.41
  14. Based on embedded image tracking system,TP391.41
  15. Exercise and defocus blurred image restoration,TP391.41
  16. Vehicle Recognition Based on Image Features,TP391.41
  17. Embedded target detection and tracking system design and algorithm implementation,TP391.41
  18. Augmented reality registration methods of virtual objects,TP391.41
  19. Target Tracking Based on Particle Filter Algorithm and DirectShow realized,TP391.41
  20. Multi-target detection and tracking method of visual and video surveillance software platform development,TP391.41
  21. Video Mean-Shift tracking algorithm applied research,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