提高SURF在小图像上的性能 [英] Improving the performance of SURF on small images

问题描述

我在网上遇到的SURF的每次实施似乎都特别糟糕，从小图片（例如，100x100或更少）中提取有用数量的兴趣点。

Every implementation of SURF I have come across on the web seems to be particularly bad at extracting a useful number of interest points from small images (say, 100x100 or less).

我尝试过多种方法：

1）使用各种升级算法（从最近邻的简单算法到更高级的算法 - 基本上每个升级版imagemagick都提供）在分析之前增加小图像的大小。

1) Using various upscaling algorithms (from simple one like nearest-neighbor to more advanced ones - basically every upscaler imagemagick provides) to increase the size of small images before analysis.

2）其他图像处理调整以显示图像中的特征，例如对比度增强和使用不同的RGB权重在积分图像的计算中。

2) Other image processing tweaks to bring out features in the image such as contrast enhancement and the use of different RGB weights in the computation of the integral image.

3）（重新）压缩，假设压缩伪像主要出现在现有特征周围，增加它们的相对表面积。

3) (Re-)compression, on the assumption that compression artifacts will appear primarily around existing features, increasing their relative "surface area."

然而，这些都没有对从小图片中提取的兴趣点数量产生任何可衡量的影响。

However, none of these has had any measurable effect on the number of interest points extracted from small images.

还有什么吗？正在尝试？或者SURF在小图像上是不好的，期间？如果是这样，那些更好的其他算法呢？

Is there anything else worth trying? Or is SURF just bad at small images, period? If so, what other algorithms are better for those?

推荐答案

这取决于你想做什么。 100x100图像不包含大量信息。此外，SURF需要为100x100图像制作有意义的描述符的描述符区域非常小。根据您的要求，尝试以下操作：

It depends what you want to do. a 100x100 image does not contain a lot of information. Also, the descriptor area that SURF needs to make a meaningful descriptor is very small for 100x100 images. Depending on what you want to do, try the following:

1. 使用整个100图像作为描述符大小。根本不检测兴趣点，而是将单个兴趣点放在100x100图像的中心（50,50）并使用整个图像数据创建描述符。这可以帮助您检测这些小图像的相似之处。使用直立SURF或方向无关SURF。


2. 使用双图像大小标记来获得更多兴趣点。


3. 减少描述符的维度使用较小的表面和较小的方块。这实际上非常适合对象识别，但不太适合3D重建。

总结一下。这一切都取决于你想要达到的目标。请随时给我留言（可能是 SURF的Github回购）

To sum it up. It all depends on what you want to achieve. Feel free to drop me a message (maybe on the Github repo of SURF)

这篇关于提高SURF在小图像上的性能的文章就介绍到这了，希望我们推荐的答案对大家有所帮助，也希望大家多多支持IT屋！