A Motion Direction Detective Mechanism for Greyscale Images

The visual system of humans and animals is considered to be the most important sensory system, and more than 80% of the necessary information in the interaction with nature comes from vision. And directional information is one of the most basic visual information. The brain needs to obtain various types of simple visual information before it can make a correct evaluation of the external environmental conditions and thus decides how to interact with the external world. Therefore, the correct detection of motion direction is also a particularly important function of the visual system. However, the exact mechanism of motion direction detection in the brain is still unknown. In this paper, we propose a motion direction detection mechanism for grayscale image to detect the direction of object motion occurring in the visual field. The main idea of this mechanism is to use multiple local motion directionally detective neurons that are sensitive to motions of different directions to collect the local motion direction information, and then infer the motion direction of object in the visual field with the global motion directionally detective neurons. A series of experimental results demonstrate the effectiveness and accuracy of the proposed mechanism, which is independent of the shape, size, and position of the object in the visual field. The mechanism is based on grayscale images, so it also has the ability to correctly detect the direction of object motion in binary images.