haar级联每个阶段的培训步骤非常多（OpenCV）

Question

我使用OpenCV训练了级联分类器，但是对于每个阶段，它都没有达到合适的误报率。培训过程的一部分如下所示：

    ===== TRAINING 7-stage =====
    <BEGIN
    POS count : consumed   13737 : 13749
    NEG count : acceptanceRatio    34344 : 0.44846
Precalculation time: 4
+----+---------+---------+
|  N |    HR   |    FA   |
+----+---------+---------+
|   1|        1|        1|
+----+---------+---------+
|   2|        1|        1|
+----+---------+---------+
|   3|        1|        1|
+----+---------+---------+
|   4|        1|        1|
+----+---------+---------+
|   5|        1|        1|
+----+---------+---------+
|   6|        1|        1|
+----+---------+---------+
|   7|        1|        1|
+----+---------+---------+
|   8|        1|        1|
+----+---------+---------+
|   9|        1|        1|
+----+---------+---------+
|  10|        1|        1|
+----+---------+---------+
|  11|        1|        1|
+----+---------+---------+
|  12|        1|        1|
+----+---------+---------+
|  13|        1| 0.999971|
+----+---------+---------+
|  14|        1| 0.999971|
+----+---------+---------+
|  15|        1| 0.999971|
+----+---------+---------+
|  16|        1| 0.999971|
+----+---------+---------+
|  17|        1| 0.999971|
+----+---------+---------+
|  18|        1| 0.999971|
+----+---------+---------+
|  19|        1| 0.999971|
+----+---------+---------+
|  20|        1| 0.999942|
+----+---------+---------+
|  21|        1| 0.999942|
+----+---------+---------+
|  22|        1| 0.999942|
+----+---------+---------+
|  23|        1| 0.999942|
+----+---------+---------+
|  24|        1| 0.999825|
+----+---------+---------+
|  25|        1| 0.999825|
+----+---------+---------+
|  26|        1| 0.999738|
+----+---------+---------+
|  27|        1| 0.999767|
+----+---------+---------+
|  28|        1| 0.999621|
+----+---------+---------+
|  29|        1| 0.999709|
+----+---------+---------+
|  30|        1| 0.999592|
+----+---------+---------+
|  31|        1| 0.999738|
+----+---------+---------+
|  32|        1| 0.999534|
+----+---------+---------+
|  33|        1| 0.999592|
+----+---------+---------+
|  34|        1| 0.999418|
+----+---------+---------+
|  35|        1| 0.999418|
+----+---------+---------+
|  36|        1| 0.999272|
+----+---------+---------+
|  37|        1|  0.99933|
+----+---------+---------+
|  38|        1| 0.999185|
+----+---------+---------+
|  39|        1| 0.999359|
+----+---------+---------+
|  40|        1| 0.998981|
+----+---------+---------+
|  41|        1| 0.999243|
+----+---------+---------+
|  42|        1| 0.998923|
+----+---------+---------+
|  43|        1|  0.99901|
+----+---------+---------+
|  44|        1|  0.99869|
+----+---------+---------+
|  45|        1| 0.998952|
+----+---------+---------+
|  46|        1| 0.998631|
+----+---------+---------+
|  47|        1| 0.998864|
+----+---------+---------+
|  48|        1| 0.998457|
+----+---------+---------+
|  49|        1| 0.998631|
+----+---------+---------+
|  50|        1| 0.997962|
+----+---------+---------+
|  51|        1| 0.998166|
+----+---------+---------+
|  52|        1| 0.997496|
+----+---------+---------+
|  53|        1| 0.997845|
+----+---------+---------+
|  54|        1| 0.997292|
+----+---------+---------+
|  55|        1| 0.997642|
+----+---------+---------+
|  56|        1| 0.996914|
+----+---------+---------+
|  57|        1| 0.997176|
+----+---------+---------+
|  58|        1| 0.996389|
+----+---------+---------+
|  59|        1| 0.996768|
+----+---------+---------+
|  60|        1| 0.995836|
+----+---------+---------+
|  61|        1| 0.996273|
+----+---------+---------+
|  62|        1| 0.995749|
+----+---------+---------+
|  63|        1|  0.99604|
+----+---------+---------+
|  64|        1| 0.995458|
+----+---------+---------+
|  65|        1| 0.995982|
+----+---------+---------+
|  66|        1|  0.99473|
+----+---------+---------+
|  67|        1| 0.995196|
+----+---------+---------+
|  68|        1| 0.994089|
+----+---------+---------+
|  69|        1| 0.994672|
+----+---------+---------+
|  70|        1| 0.993216|
+----+---------+---------+
|  71|        1| 0.994089|
+----+---------+---------+
|  72|        1| 0.992633|
+----+---------+---------+
|  73|        1| 0.993565|
+----+---------+---------+
|  74|        1| 0.991789|
+----+---------+---------+
|  75|        1| 0.992866|
+----+---------+---------+
|  76|        1| 0.991119|
+----+---------+---------+
|  77|        1| 0.991964|
+----+---------+---------+
|  78|        1| 0.990217|
+----+---------+---------+
|  79|        1| 0.991673|
+----+---------+---------+
|  80|        1| 0.989867|
+----+---------+---------+
|  81|        1| 0.990886|
+----+---------+---------+
|  82|        1| 0.988178|
+----+---------+---------+
|  83|        1| 0.989372|
+----+---------+---------+
|  84|        1| 0.987218|
+----+---------+---------+
|  85|        1| 0.988499|
+----+---------+---------+
|  86|        1| 0.985936|
+----+---------+---------+
|  87|        1| 0.987509|
+----+---------+---------+
|  88|        1| 0.984888|
+----+---------+---------+
|  89|        1|  0.98681|
+----+---------+---------+
|  90|        1|  0.98352|
+----+---------+---------+
|  91|        1| 0.985762|
+----+---------+---------+
|  92|        1| 0.981918|
+----+---------+---------+
|  93|        1|  0.98451|
+----+---------+---------+
|  94|        1| 0.980259|
+----+---------+---------+
|  95|        1| 0.982879|
+----+---------+---------+
|  96|        1| 0.978308|
+----+---------+---------+
|  97|        1| 0.981249|
+----+---------+---------+
|  98|        1| 0.976532|
+----+---------+---------+
|  99|        1| 0.979647|
+----+---------+---------+
| 100|        1| 0.975862|
+----+---------+---------+
END>
Training until now has taken 0 days 0 hours 8 minutes 41 seconds.

大约2000个样本是真实世界的样本，其他样本是由opencv_createsample工具创建的。我不确切地知道问题是什么。我想增强图像很难辨别，是真的吗？