Abstract
This paper proposes a new tiered heterogeneous Wireless image sensor network for real-time unobtrusive species detection and video cataloguing based on a new HIerarchically Scalarized - Character Oriented DEtection (MS-CODE) algorithm and architecture. An experimented test bed has been set up at the Omaha Henry Doorly Zoo, and data collected over wireless network from this test bed has been processed and classified based on the proposed HIS-CODE. This work contributes to a novel integration in wireless sensor networks with the in-network distributed image recognition functionality, especially with the consideration of wildlife species image recognitions.