Skip to Main Content
Notice of Retraction
After careful and considered review of the content of this paper by a duly constituted expert committee, this paper has been found to be in violation of IEEE's Publication Principles.
We hereby retract the content of this paper. Reasonable effort should be made to remove all past references to this paper.
The presenting author of this paper has the option to appeal this decision by contacting TPII@ieee.org.
Image feature and similarity measure are important topics in content-based image retrieval. In this paper, we present energy signal sequences of Energy entropy, Entropy, Averagy residual, Standard deviation from Pulse-Coupled Neural Networks (PCNN) as image feature respectively, and Correlation Coefficient (CC) as the similarity metrics in image retrieval system. The pulse image sequence generated by PCNN contain a large amount of original image information, and are invariant to translation, rotation, scaling and distortion, and they can be calculated to the energy signal sequence as the image feture. CC is an excellent criteria in comparison of image similarity, which has an inherent ability to suppress noise and is robust to image rotation and scaling. The experimental results show that the Averagy residual signal sequence feature outperforms of the other three features with comprehensive consideration, and the Correlation coefficient based retrieval system has much better geometric invariance and stronger antinoise ability than the traditional Euclidean Distance (ED) based system.