Journal of Kunming Metallurgy College ›› 2021, Vol. 37 ›› Issue (1): 86-89.DOI: 10.3969/j.issn.1009-0479.2021.01.016
Previous Articles Next Articles
SHI Hongliang
Received:
Online:
Published:
Abstract: In order to reduce the number of model parameters and iteration times of the Pulse Coupled Neural Network (PCNN) in image segmentation,an improved PCNN fast image segmentation algorithm is proposed. PCNN is simplified and the constant connection coefficient is linked with pixel gray scale value of corresponding neurons in the algorithm. The algorithm does not need to set PCNN parameters artificially in image segmentation and the dynamic threshold is transformed into constant threshold according to gray scale statistical characteristics of the image so that only one iteration time is needed for image segmentation. 'The experimental results show that the proposed algorithm has good performance of subjective visual perception and lower time complexity,and it is better than comparison algorithm.
Key words: image segmentation, parameter setting, PCNN model improved, iteration times
CLC Number:
TP911.73
SHI Hongliang. An Improved Image Segmentation Algorithm Based on PCNN[J]. Journal of Kunming Metallurgy College, 2021, 37(1): 86-89.
0 / / Recommend
Add to citation manager EndNote|Ris|BibTeX
URL: http://kmyzxb.magtech.com.cn/EN/10.3969/j.issn.1009-0479.2021.01.016
http://kmyzxb.magtech.com.cn/EN/Y2021/V37/I1/86