Dissertation > Excellent graduate degree dissertation topics show
The Research on Algorithms for Recognition of Cow Image Based on Wavelet Moment
Author: JinYiChu
Tutor: MaYan
School: Shanghai Normal University
Course: Applied Computer Technology
Keywords: Cows recognition Homomorphic filtering Regional Growth Eight-connected component labeling Wavelet Moment
CLC: TP391.41
Type: Master's thesis
Year: 2011
Downloads: 17
Quote: 0
Read: Download Dissertation
Abstract
|
The cows identification is an important part of the management of the cows , and the previous identify cows mainly rely on manual identification , mainly uses electronic tags to identify cows picture , and use the computer to identify cows management is a new field . Cows image recognition is divided into three steps : to remove the impact of light , image segmentation and object recognition . In this paper, based on the homomorphic filtering processing algorithms wavelet transform and region growing algorithm based on the sub-block in the YCbCr color space , while eight neighborhood marking algorithm improvements . This paper completed the following work : (a) to remove the impact of light on the image segmentation and recognition . The light shines in the cows body surface reflective , so that part of the area of the cow in the picture , whose colors and the color is not the same as the original . In this paper, the YCbCr color space the homomorphic filtering processing algorithm based on wavelet transform . The image is first converted to the YCbCr color space, the luminance component using wavelet decomposition to filter unnecessary detail and with homomorphic filter removing light . (B ) In order to improve recognition speed , region growing algorithm based on sub-block - based region growing algorithm . 4 × 4 sub-block as a unit, as a sub-block is characterized first block inside color clustering and extract its color and texture features , and then according to the criteria subblock region growing . After the initial segmentation , this paper Finally, we propose an improved eight - neighborhood labeling algorithm to repair cracks and small holes within the target segmentation segmentation algorithm based on Otsu multi- threshold separating the target from the background . ( C) split the image to select the area of ??the largest black markings as target identification . First goal such as pan and zoom normalization process , followed with a three-dimensional B - spline function wavelet moment , and finally calculate the wavelet moment extracted eigenvalues ??with the eigenvalues ??of the cows in the database to compare obtained recognition results . Experiments show that the recognition algorithm proposed in this paper can effectively identify cows , the identification rate of about 84% .
|
Related Dissertations
- Research on Region of Interest Based Image Retrieval,TP391.3
- Extraction of planar target in remote sensing image based on mathematical morphology,P237
- Study on Gait-based Human Identification,TP391.4
- Control System and Control Method on Intelligent Wire Bonder,TN405.96
- A Study of Coastline Extraction in RS Image and It’s Change Detection Technology,P237
- Research of the Methods in Removing Cloud of Remote Sensing Image,TP751
- Design and Implementation of Human Activity Recognition System Based on Wavelet Moment,TP391.41
- Research on Classification and Matching for Hand Vein Image,TP391.41
- The Research of Infrared Image Preprocessing and Small Target Detection under Complex Background,TP391.41
- Research on Red Tide Algal Cell’s Recognition,TP391.4
- Research on the Cloud Removal Method of High Resolution Remote Sensing Images,TP751
- Research of Moving Object Detection and Tracking Technology and DSP Implementation,TP274.4
- 3D segmentation of medical CT image and Visualization,R318
- Study of the Interactive Image Segmentation Algorithms Based on Fuzzy Connectedness,TN911.73
- License Plate Locating System in Complex Environment,TP391.41
- The Technology of Interactive Virtual Hepatic Surgery System,TP391.41
- The Segmentation and Statistic for Human Chromosome Image,TP391.41
- Research on Digital Watermarking of Contourlet Domain,TP309.7
- Processing and Recognition of Defects in X-ray Tire Image,TP391.41
- Algorithm Study on Phase Extraction and Phase Unwrapping in Phase Measuring Profilometry Based on Projection,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
|