红瓤西瓜瓤色数字化分类方法初探

The method of digital classification of red flesh watermelon flesh colors

随着电子信息技术的发展,西瓜瓤色的数字图像被大量保存使用,为了实现西瓜瓤色的数字化分类,笔者以粉红、桃红、橘红、大红瓤色西瓜材料的数字图像为研究对象,对比了不同光照条件下西瓜瓤色的数码图像,以粉红、桃红、红、橘红、大红瓤色西瓜材料的数字图像为研究对象,测量了西瓜瓤色的数码图像在RGB模式下的数值,转换为HSB模式后,通过比较L×S值进而比较西瓜瓤色。结果表明,西瓜瓤色在室外阴、日光灯和阳光3种光照条件下拍摄的数字图像差别较小,在闪光灯、室内、暗室的光照条件下拍摄的数字图像颜色变化较明显。粉红、桃红、红、橘红、大红瓤色的L×S值变化区间分别为22.6~55.0、46.7~65.9、62.3~75.2、68.8~78.4、74.4~81.7,初步实现了西瓜瓤色的数字化分类。

With the development of electronic information technology, a large number of digital images of watermelon flesh color have been preserved. In order to realize the digital elassification of watermelon flesh colors, the digital images of watermelon flesh colors （pink, peach, orange and bright red） were used as materials, the changes of these digital images in different light conditions were compared, the value of these digital images under the RGB model were measured, after converted to ItSB mode, LxS value was used to differentiate watermelon flesh colors. The results showed that there was little difference in digital images of watermelon flesh colors carried out by outdoor shadow, daylight lamp and sunlight, obvious difference was found in the light condition of flashlight, indoor light （no supplementary light） and darkroom. The variation ranges of flesh colors of pink, peach, orange and bright red were 22.6-55.0, 46.7-65.9, 62.3-75.2, 68.8-78.4, 74.4-81.7 respectively, the digital classification of watermelon flesh color was initially realized.