Moving object detection based on bioinspired background subtraction

• DOI: 10.1088/1748-3190/ad5ba3
• Publication Date: 2024-06-25T22:41:26Z
• Authors: Zhu’anzhen Zheng; Aike Guo; Zhihua Wu
• Keywords: Motion Detection; Subtraction

Flying insects rely mainly upon visual motion to detect and track objects. There has been a lot of research on fly inspired algorithms for object detection, but few have developed based alone. One the daunting difficulties is that neural circuit mechanisms underlying foreground-background segmentation are still unclear. Our previous modeling study proposed lobula held parallel pathways with distinct directional selectivity, each which could retinotopically discriminate figures moving in its own preferred direction relative cues. The model, however, did not address how multiple gave only detection output at their common downstream. Since directions along either horizontal or vertical axis were opposite other, background an also activated corresponding pathway. Indiscriminate ungated projection from all downstream would mix objects background, making model fail non-stationary background. Here, we extend by proposing motion-dependent gating individual projections key detection. Large-field plate tangential cells hypothesized perform realize bioinspired subtraction. shown be capable implementing robust video sequences camera induces translational optic flow static camera. sheds light potential concise algorithm real-world applications.