Improved recognition of Echinococcus granulosus protoscoleces using visual saliency and scale-invariant features
1 School of Electronics and Information Engineering, Harbin Institute of Technology, Harbin 150001, China
2 Xinjiang Key Laboratory of Echinococcosis, Clinical Medical Research Institute, The First Affiliated Hospital of Xinjiang Medical University, Urumqi 830054, China
3 College of Information Science and Engineering, Xinjiang University, Urumqi 830046, China
Figures & Tables
Fig. 1. (a) Microscopic image of Echinococcusgranulosus protoscoleces and (b) the corresponding visual saliency map produced with the Itti model.
下载图片 查看原文
Fig. 2. Graphical depiction of the difference-of-Gaussian function.
下载图片 查看原文
Fig. 3. Keypoint descriptor determination process.
下载图片 查看原文
Fig. 4. Proposed method for detecting Echinococcusgranulosus protoscoleces based on visual saliency and scale-invariant features.
下载图片 查看原文
Fig. 5. (a) Microscopic image of Echinococcusgranulosus protoscoleces, (b) the corresponding visual saliency map produced with the modified Itti model, and (c) the SVM classification results applied to the original image.
下载图片 查看原文
Fig. 6. Egg recognition results using the proposed method.
下载图片 查看原文
Table1. Comparative Parasite Recognition Rates for Methods Using the Conventional Itti model and the proposed modified Itti model
Visual Saliency Model | Number of Actual Living Parasites | Number of Correctly Identified Living Parasites | Parasite Recognition Rate |
---|
Conventional Itti model | 22 | 18 | 81.8% | Modified Itti model | 22 | 21 | 95.4% |
|
查看原文
Table2. Comparative Parasite Recognition Rates Using the Proposed Method for Three Different Eosin Exclusion Test Images Containing Living Parasites
Test Image | Number of Actual Living Parasites | Number of Correctly Identified Living Parasites | Egg Recognition Rate |
---|
| 22 | 21 | 95.4% | | 30 | 27 | 90.0% | | 50 | 46 | 92.0% |
|
查看原文
Zhuang Li, Guodong Lü, Xiaoyi Lü. Improved recognition of Echinococcus granulosus protoscoleces using visual saliency and scale-invariant features[J]. Chinese Optics Letters, 2019, 17(4): 041703.