Rain sensor using statistical analysis

Abstract

A method and an assembly for sensing moisture (20) on the exterior surface (22) of a sheet of glass (14) by positioning an imaging lens (16) adjacent to the interior surface (18) of the sheet of glass (14), passing light rays from moisture (20) on the exterior surface (22) of the sheet of glass (14) through the imaging lens (16) and producing first and second successive images of the moisture (20) on the glass (14), and successively directing the first and second images from the lens (16) onto a focal plane detector (12). The focal plane detector (12) is divided into rows of pixels (24) which are ligned into columns. The assembly and method are characterized by scanning the pixels (24) of each row in a first scan (FIG. 4) and producing a reference value X for each pixel (24) in response to an initial image in that pixel (24), scanning the pixels (24) of each row in a second scan (FIG. 5) and producing a subsequent value Y different from the reference value in response to an subsequent image in that pixel (24). A correlation coefficient value r is developed for each row with a statistical formula from the reference and subsequent values to produce a wiper signal in response to the correlation coefficient value r deviating from a predetermined value. An additional condition to the wiper signal could be that the correlation coefficient values r differ from row to row.

Dependencies are inferred from the claim text.