另一个不太可靠的,但可能更快解决方案是构建特征直方图的每个图像,并且选择图像与最接近直方图于输入图像的直方图。
Another less robust but potentially faster solution is to build feature histograms for each image, and choose the image with the histogram closest to the input image's histogram.
直方图特征不受目标的外形和比例变化的影响,具有较好的稳定性。
Histogram character is not affected by object figure's variance and scale, and it has good stability.
然后利用直方图法对每一特征作了特征选择,如此共选取了21个特征。
Then trait selection has been for every trait using the histogram method and 21 traits in all have been selected.
应用推荐