Progressive JPEG for Reliable Visual Monitoring in Agricultural IoT Over LoRaWAN

In data driven agriculture, continuous monitoring of crop conditions and environmental variables is essential for improving productivity and supporting informed decision making. Low Power Wide Area Network (LPWAN) protocols are well suited for monitoring large agricultural landscapes, while camera based Internet of Things (IoT) systems provide more detailed and accurate data than scalar sensors. However, the inherent limitations of IoT networks including low power availability, restricted bandwidth, and frequent packet loss make reliable image transmission difficult. This study presents an experimental evaluation of Progressive Joint Photographic Experts Group (P-JPEG) and Baseline Joint Photographic Experts Group (B-JPEG) for image transmission over Long Range Wide Area Network (LoRaWAN). P-JPEG encodes images in multiple refinement scans, enabling partial reconstruction and preserving visual content even under 50% packet loss, whereas B-JPEG images become corrupted with as little as 5% loss. Image transfer experiments over an 8 km LoRaWAN link show that P-JPEG achieves an average Structural Similarity Index Measure (SSIM) of 0.92 and a Peak Signal-to-Noise Ratio (PSNR) of 30 dB with 15% packet loss, significantly outperforming B-JPEG. Additional field tests conducted in an agricultural environment, with a 0.98 km separation between the node and the gateway, demonstrate that a 2 kB image can be transmitted in 30.6 seconds, achieving a data rate of 522.8 bps. Overall, the results show that$\mathbf{P}$-JPEG offers a robust, bandwidth-efficient, and reliable solution for image transmission, making it highly suitable for visual monitoring in low-power agricultural IoT systems.