Data Augmentation for Small Sample Iris Image Based on a Modified Sparrow Search Algorithm
Abstract Training convolutional neural networks (CNN) often require a large amount of data. However, for some biometric data, such as fingerprints and iris, it is often difficult to obtain a large amount of data due to privacy issues. Therefore, training the CNN model often suffers from specific pro...
| Main Authors: | , , , |
|---|---|
| Format: | Article |
| Language: | English |
| Published: |
Springer
2022-12-01
|
| Series: | International Journal of Computational Intelligence Systems |
| Subjects: | |
| Online Access: | https://doi.org/10.1007/s44196-022-00173-7 |
| _version_ | 1797980479615401984 |
|---|---|
| author | Qi Xiong Xinman Zhang Shaobo He Jun Shen |
| author_facet | Qi Xiong Xinman Zhang Shaobo He Jun Shen |
| author_sort | Qi Xiong |
| collection | DOAJ |
| description | Abstract Training convolutional neural networks (CNN) often require a large amount of data. However, for some biometric data, such as fingerprints and iris, it is often difficult to obtain a large amount of data due to privacy issues. Therefore, training the CNN model often suffers from specific problems, such as overfitting, low accuracy, poor generalization ability, etc. To solve them, we propose a novel image augmentation algorithm for small sample iris image in this article. It is based on a modified sparrow search algorithm (SSA) called chaotic Pareto sparrow search algorithm (CPSSA), combined with contrast limited adaptive histogram equalization (CLAHE). The CPSSA is used to search for a group of clipping limit values. Then a set of iris images that satisfies the constraint condition is produced by CLAHE. In the fitness function, cosine similarity is used to ensure that the generated images are in the same class as the original one. We select 200 categories of iris images from the CASIA-Iris-Thousand dataset and test the proposed augmentation method on four CNN models. The experimental results show that, compared with the some standard image augmentation methods such as flipping, mirroring and clipping, the accuracy and Equal Error Rate (EER)of the proposed method have been significantly improved. The accuracy and EER of the CNN models with the best recognition performance can reach 95.5 and 0.6809 respectively. This fully shows that the data augmentation method proposed in this paper is effective and quite simple to implement. |
| first_indexed | 2024-04-11T05:55:04Z |
| format | Article |
| id | doaj.art-1ef3c79ed6d844b796f980d7d1b1eab3 |
| institution | Directory Open Access Journal |
| issn | 1875-6883 |
| language | English |
| last_indexed | 2024-04-11T05:55:04Z |
| publishDate | 2022-12-01 |
| publisher | Springer |
| record_format | Article |
| series | International Journal of Computational Intelligence Systems |
| spelling | doaj.art-1ef3c79ed6d844b796f980d7d1b1eab32022-12-22T04:41:55ZengSpringerInternational Journal of Computational Intelligence Systems1875-68832022-12-0115111110.1007/s44196-022-00173-7Data Augmentation for Small Sample Iris Image Based on a Modified Sparrow Search AlgorithmQi Xiong0Xinman Zhang1Shaobo He2Jun Shen3International Collage, Hunan University of Arts and SciencesSchool of Automation Science and Engineering, Faculty of Electronic and Information Engineering, MOE Key Lab for Intelligent Networks and Network Security, Xian Jiaotong UniversitySchool of Physics and Electronics, Central South UniversitySchool of Computing and Information Technology, University of WollongongAbstract Training convolutional neural networks (CNN) often require a large amount of data. However, for some biometric data, such as fingerprints and iris, it is often difficult to obtain a large amount of data due to privacy issues. Therefore, training the CNN model often suffers from specific problems, such as overfitting, low accuracy, poor generalization ability, etc. To solve them, we propose a novel image augmentation algorithm for small sample iris image in this article. It is based on a modified sparrow search algorithm (SSA) called chaotic Pareto sparrow search algorithm (CPSSA), combined with contrast limited adaptive histogram equalization (CLAHE). The CPSSA is used to search for a group of clipping limit values. Then a set of iris images that satisfies the constraint condition is produced by CLAHE. In the fitness function, cosine similarity is used to ensure that the generated images are in the same class as the original one. We select 200 categories of iris images from the CASIA-Iris-Thousand dataset and test the proposed augmentation method on four CNN models. The experimental results show that, compared with the some standard image augmentation methods such as flipping, mirroring and clipping, the accuracy and Equal Error Rate (EER)of the proposed method have been significantly improved. The accuracy and EER of the CNN models with the best recognition performance can reach 95.5 and 0.6809 respectively. This fully shows that the data augmentation method proposed in this paper is effective and quite simple to implement.https://doi.org/10.1007/s44196-022-00173-7Data augmentationSmall sampleSwarm intelligenceSparrow search algorithmIris images |
| spellingShingle | Qi Xiong Xinman Zhang Shaobo He Jun Shen Data Augmentation for Small Sample Iris Image Based on a Modified Sparrow Search Algorithm International Journal of Computational Intelligence Systems Data augmentation Small sample Swarm intelligence Sparrow search algorithm Iris images |
| title | Data Augmentation for Small Sample Iris Image Based on a Modified Sparrow Search Algorithm |
| title_full | Data Augmentation for Small Sample Iris Image Based on a Modified Sparrow Search Algorithm |
| title_fullStr | Data Augmentation for Small Sample Iris Image Based on a Modified Sparrow Search Algorithm |
| title_full_unstemmed | Data Augmentation for Small Sample Iris Image Based on a Modified Sparrow Search Algorithm |
| title_short | Data Augmentation for Small Sample Iris Image Based on a Modified Sparrow Search Algorithm |
| title_sort | data augmentation for small sample iris image based on a modified sparrow search algorithm |
| topic | Data augmentation Small sample Swarm intelligence Sparrow search algorithm Iris images |
| url | https://doi.org/10.1007/s44196-022-00173-7 |
| work_keys_str_mv | AT qixiong dataaugmentationforsmallsampleirisimagebasedonamodifiedsparrowsearchalgorithm AT xinmanzhang dataaugmentationforsmallsampleirisimagebasedonamodifiedsparrowsearchalgorithm AT shaobohe dataaugmentationforsmallsampleirisimagebasedonamodifiedsparrowsearchalgorithm AT junshen dataaugmentationforsmallsampleirisimagebasedonamodifiedsparrowsearchalgorithm |