| 摘要 |
PURPOSE: An automatic reading method of weather condition from image data, and an image processing method thereof are provided to read weather condition automatically from image data, and to change a color different from a representative color tone, extracted from the entire region of the read image to the complementary color of the representative color tone. CONSTITUTION: An image processor separates signals of digital image into a luminance signal and a chrominance signal. The image processor obtains a standard deviation of a histogram of the luminance signal. I case the standard deviation of the histogram is smaller than a reference standard deviation, the image processor determines the digital image as a weather condition that a natural phenomenon is occurred, the natural phenomenon is one kind selected from a group comprising fog, yellow dust, and smog. [Reference numerals] (AA) Analog video input; (BB) Video A/V conversion; (CC) VC image separation; (DD) Image correction; (EE) Histogram treatment; (FF) Histogram average value extraction; (GG) Average value > reference value?; (HH) Y: daytime; (II) N: night; (JJ) Histogram standard deviation distribution area extraction; (KK) Standard deviation < reference deviation; (LL) Y: fog, yellow sand, smog; (MM) Representative color=fog?, representative color=yellow sand?, representative color=smog?, representative color=snow, rain?; (NN) Contract reinforcement; (OO) Highpass filter; (PP) Median filter; (QQ) Original image=median process result image; (RR) N: snow, rain, hail, and the like; (SS) Set the complementary color of a representative color |
