They are also widely used in aerial platforms for obstacle detection, stabilisation, and navigation. They are very popular in autonomous vehicles for detecting and tracking pedestrians or other vehicles. The ratio of these two values makes it possible to determine the direction of the movement, while their magnitude allows to determine the speed.Īlgorithms that perform this type of task are increasingly used in everyday life. Hence, the concept of optical flow was introduced, i.e., a vector field describing the movement of a pixel between two images from a sequence, in which two values are associated with each pixel-its horizontal and vertical displacements. However, this information is essential in more advanced vision systems. This allows to see where the movement (change) has occurred, but does not provide significant information about its direction and speed. In the simplest case, the motion can be detected by subtracting two subsequent frames from a video sequence. The algorithms realised in this work can be a component of a larger vision system in advanced surveillance systems or autonomous vehicles.Įxemplary frames from Camera motion sequence. The presented solution allows real-time optical flow determination in multiple scales for a 4K resolution with estimated energy consumption below 6 W. Depending on the scale, the calculations were performed for different data formats, allowing for more efficient processing by reducing resource utilisation. In order to detect larger pixel displacements, a multi-scale approach was used in both algorithms. A vector data format was used to enable flow calculation for a 4K (Ultra HD, 3840 × 2160 pixels) video stream at 60 fps. In this work, two gradient-based algorithms-Lucas–Kanade and Horn–Schunck-were implemented on a ZCU 104 platform with Xilinx Zynq UltraScale+ MPSoC FPGA. This represents a challenging task, especially for high-resolution video streams. In some robotic applications, e.g., in autonomous vehicles, it is necessary to calculate the flow in real time. The information about optical flow, i.e., the movement of pixels between two consecutive images from a video sequence, is used in many vision systems, both classical and those based on deep neural networks.
0 Comments
Leave a Reply. |
AuthorWrite something about yourself. No need to be fancy, just an overview. ArchivesCategories |