A Signal Contour Extraction Algorithm based on Canny Operator and its Application in Bionic Communication
- Resource Type
- Conference
- Authors
- Huang, Xiangdang; Zhang, Jiangyi; Li, Pengcheng; Chan, Shihao; Lian, Kaiwei; Chen, Yanxia
- Source
- 2023 IEEE 29th International Conference on Parallel and Distributed Systems (ICPADS) ICPADS Parallel and Distributed Systems (ICPADS), 2023 IEEE 29th International Conference on. :2698-2705 Dec, 2023
- Subject
- Communication, Networking and Broadcast Technologies
Computing and Processing
Time-frequency analysis
Image edge detection
Simulation
Merging
Filtering algorithms
Feature extraction
Frequency estimation
Covert hydroacoustic communication
Dolphin Whistle bionic communication
Canny edge detection
Contour extraction
- Language
- ISSN
- 2690-5965
Bionic communication offers advantages over traditional covert hydroacoustic communication in terms of communication distance and security. However, there is still room for further development in the area of sound signal feature extraction. In this paper, we propose a signal contour extraction algorithm with the dolphin whistle as the research subject: the Edge Matrix Merging Extraction Method based on the Canny algorithm (EMMEC). EMMEC better utilizes the structural characteristics of the signal's time-frequency spectrum and enhances the accuracy of whistle time-frequency curve extraction compared to traditional methods. EMMEC initially consolidates and processes whistle information using edge detection to generate a merged edge matrix. It then extracts the time spectrum profile curve of dolphin whistles by estimating curve parameters. Simulation experiments demonstrate that our proposed method can more accurately and smoothly fit the whistle time-frequency spectrum curve by effectively leveraging the structural characteristics of the signal's time-frequency spectrum, ultimately improving the accuracy of curve extraction.