A Novel Hierarchical Coding Progressive Transmission Method for WMSN Wildlife Images

Area of Science:

• Computer Science
• Signal Processing
• Wildlife Monitoring

Background:

• Wireless multimedia sensor networks (WMSNs) face bandwidth limitations and signal interference, hindering reliable wildlife image transmission.
• Transmitting entire wildlife images is inefficient as only specific regions (e.g., animals) hold critical information.
• Existing methods struggle with data loss and transmission failures in real-world wildlife monitoring scenarios.

Purpose of the Study:

• To develop an efficient hierarchical coding progressive transmission method for wildlife images.
• To improve the quality and reliability of wildlife image transmission in WMSNs.
• To address bandwidth constraints and signal interference issues in wildlife monitoring.

Main Methods:

• A convolution neural network using MobileNet was employed for saliency object detection and mask generation of wildlife.
• Set Partitioned In Hierarchical Tree (SPIHT) lossless coding was used for transmitting the important saliency regions.
• Embedded Zerotree Wavelets (EZW) lossy coding was applied to transmit the background regions, optimizing efficiency.

Main Results:

• The proposed method successfully transmitted field-captured wildlife images with improved quality.
• Demonstrated significant enhancements in Peak Signal to Noise Ratio (PSNR) and Structural Similarity Index (SSIM) compared to existing algorithms.
• Achieved PSNR and SSIM improvements of 21.11%, 14.72% and 9.47%, 6.25% over EZW and DCT methods, respectively.

Conclusions:

• The hierarchical coding progressive transmission method effectively addresses challenges in WMSN wildlife monitoring.
• This approach balances transmission efficiency with high-quality image reconstruction, prioritizing critical visual information.
• The method offers a robust solution for reliable and efficient wildlife image data acquisition and transmission.