A Fidelity Evaluation Method for Watermarked Vector Maps Based on AHP-Fuzzy Comprehensive Evaluation and BP Neural Network
Digital watermarking,a cutting-edge technology for vector map security,can impact the quality of the maps to a certain extent.The existing research usually assesses watermarked vector maps'fidelity and invisibility via visual observation and error analysis,often ignoring the maps'intrinsic features,resulting in imprecise evaluations.Addressing these issues,this paper proposes a fidelity evaluation method for watermarked vector maps based on the analytic hierarchy process(AHP)fuzzy comprehensive evaluation and back propagation(BP)neural network.The comprehensive evaluation model considers key dimensions such as topological quality,geometric features,and coordinate errors.It enhances evaluation accuracy and objectivity by optimizing subjective and objective weights through a combination of fuzzy comprehensive evaluation and BP neural network.The experimental results indicate that the proposed method provides fidelity evaluation results for watermarked vector maps that more accurately reflect the actual data,offering a comprehensive and objective framework for evaluating digital watermarking algorithms,which is more scientifically robust and rational than conventional error analysis approaches.Furthermore,this method is significant in determining the optimal watermark embedding strength,and can guide the adjustment of watermark algorithm parameters according to different application scenarios and data characteristics,providing a basis for balancing the comprehensive performance of watermarking algorithms.
digital watermarkingvector mapsfidelity evaluationfuzzy comprehensive evaluationneural network
万方数据
维普
