Downward-view pedestrian risk quantification method based on two-dimensional world coordinate system

[0101] Example 1:

[0102] For the vehicle video of the measured road scene with a pixel size of 1920×1080, this embodiment adopts the patent to quantify the adjacent pedestrian risk coefficients of two pedestrian targets in the image. The calculation results of the risk coefficient of adjacent pedestrians can be seen Figure 11 In (a), (b), (c) and (d), it can be seen that for the two pedestrian targets in the image crossing the front area of ​​the vehicle, reasonable pedestrian risk quantification results are given.

[0103] Example 2:

[0104] In this example, for the two pedestrian targets in the vehicle video of the measured road scene with a size of 1920×1080, the calculation results of the risk coefficient of adjacent pedestrians are given as follows: Figure 12 shown in (a), (b), (c) and (d). It can be seen that the present invention provides accurate pedestrian risk quantification results for pedestrian targets that are not related to vehicles.

[0105] Example 3:

[0106] In this example, the vehicle video is a measured road scene image with a pixel size of 1920 × 1080, and 2 pedestrian targets in the continuous image are quantified. The calculation results of the adjacent pedestrian risk coefficient are as follows: Figure 13 (a), (b), (c) and (d). It can be seen that, for the pedestrian crossing the area in front of the vehicle in the video image, the present invention provides accurate pedestrian risk quantification results.