Study on adaptation and focusing in the visual system and its hardware implementation.

(1)利用马赫带的出现和消失机制进行焦域边缘检测以前，人们认为马赫带起到增强模糊边缘的作用。Tsukada等人。模型模拟表明，在视觉系统中，马赫带的消失起着重要作用，而不是马赫带的出现，并且有助于进行焦点边缘检测。为了验证这一建议，我们在真实图像中模拟了边缘检测。结果，在摄像机距离环设置为50 cm的条件下，马赫带在48 cm到52 cm的有限范围内消失。此外，我们还建立了上述软件模型的二维硬件模型，并对马赫带的出现和消失进行了模拟。(2)视网膜视锥适应的硬件模型虽然视锥对光的动态范围最多为3个对数单位，但在大约7个对数单位的亮度范围内，视锥细胞可以在视觉上对光做出响应。圆锥体通过在时空上相对于环境光级别改变适应级别来实现这一点。我们提出了一个锥体自适应的数学模型。但我们没有考虑到空间特征。在这项研究中，我们开发了一个硬件模型，并试图通过考虑周长自适应来模拟视锥的空间和时间特性。

(1) Focal edge detection using mechanism of appearance and disappearance of Mach bandsPreviously, it is considered that Mach bands perform to enhance a blur edge. Tsukada et al. suggested by model simulation that the disappearance of Mach bands plays an important role rather than the appearance of them and contributes to a focal edge detection in a visual system. To verify this suggestion, we simulated the edge detection in a real-world image. As a result, Mach bands disappeared in a limited range from 48cm to 52cm under the condition that the distance ring of the camera was set to 50cm. Furthermore, we developed a two dimensional hardware model of above software model and simulated the appearance and disappearance of Mach bands.(2) Hardware model on adaptation of retinal coneAlthough the dynamic range of retinal cones to the light is three log units at most, the cones can visually respond to the light over a luminance span of about seven log units. The cones accomplish this by shifting the adapting level relative to the ambient light levels spatio-temporally. We proposed a mathematical model on a cone adaptation. But we did not consider the spatial characteristics. In this research, we developed a hardware model and attempted to simulate the spatial and temporal characteristics of the cones by considering circumference adaptation.