| 摘要 |
The invention relates to a method for sensing and measuring vegetation in the spatial surroundings of traffic routes while using a sensor unit that combines the reflectivity of the objects located in this spatial area in the infrared and visible spectral region and/or in the infrared and ultraviolet spectral region. In doing this, the presence of vegetation is not only detected but this existing vegetation is quantifiably sensed, in a defined manner and in geometric dimensions, reliably with regard to other useful objects present in the immediate vicinity of the traffic route. To this end, the invention provides that a one-dimensional or two-dimensional difference image of the spatial area sensed by the sensor unit is generated by means of pixel-by-pixel subtraction between the respective gray scale value intensities of a sensor unit-recorded image of the visible or ultraviolet color channel and of a sensor unit-recorded image of the infrared color channel. A quantity of one-dimensional or two-dimensional pixels that characterizes the vegetation region is extracted by means of a threshold value analysis of the pixel points of this difference image. In addition: a subset of pixel points of the image recorded by the sensor unit, this subset containing at least one pixel point assigned to the vegetation region, is selected; the pixel area with the highest correlation to this subset is determined while using known correlation-based and/or characteristic-based pattern comparison methods in an image of the same spatial area, this image being record by a second camera, and; the removal of this at least one pixel point, which is to be assigned to the vegetation region, from the plane extending from the sensor unit and second camera is determined using known stereogrammetric methods. |
