摘要
In recent years, the pulse coupled neural network (PCNN) is widely used in image segmentation, but the existing algorithm has an unsatisfactory performance with the definition of single threshold, especially in uneven illumination or low contrast image. In this paper, in order to eliminate the impact of the illumination and improve the adaptability, deblocking PCNN algorithm is utilized. It segments the image into several rectangles with the same size. Then PCNN algorithm is applied to segment each block and stopped by improved OSTU. The emulation experiment shows that this method is better than traditional image segmentation method in uneven illumination or low contrast image.
延伸閱讀
- 謝克瑤（2004）。A Block-based Approach for Partial Document Image Matching〔碩士論文，國立臺灣師範大學〕。華藝線上圖書館。https://www.airitilibrary.com/Article/Detail?DocID=U0021-2004200710592067
- Tsai, C. S., Wu, H. C., Lee, C. C., Shie, T. F., & Lee, C. C. (2017). A Steganographic Method by Pixel-Value Differencing and Exploiting Modification Direction. 電腦學刊, 28(1), 13-29. https://doi.org/10.3966/199115592017022801002
- Chou, P. H. (2014). A Semi-Distributed Reputation Mechanism based on Dynamic Data-Driven Application System [master's thesis, National Chiao Tung University]. Airiti Library. https://doi.org/10.6842/NCTU.2014.00171
- Chen, C. S., Wang, S. R., & Liu, T. C. (2007). DNS-based Network Anomaly Detection and Eradicating Scheme. 網際網路技術學刊, 8(3), 329-335. https://www.airitilibrary.com/Article/Detail?DocID=16079264-200707-8-3-329-335-a
- 施文富（2007）。An Adaptive Anomaly Detection Method Based on Incremental Hidden Markov Model and Windows Native API〔碩士論文，國立中央大學〕。華藝線上圖書館。https://www.airitilibrary.com/Article/Detail?DocID=U0031-0207200917345047