我正在尝试将available generalized hough transform (GHT)应用于我自己的数据。程序在提供的示例数据上运行得非常好，但是，对于我的数据，一旦达到this line，它就会出错： 我的数据已保存到main function中的两个numpy数组中：
f = h5py.File(img_path,'r') # reading the reference image
refim = f['image'].value
refim = np.asarray(refim)
refim[refim!=1]=0
#im = imread('Input1.png')
f = h5py.File(im_path,'r') # reading the image that should be matched
im = f['image'].value
im = np.asarray(im)
参考图像和测试图像的大小256x256相同，参考图像中的对象中心为[ 83.02902047 127.19376853]。名称为table的变量是形状为(90,)的列表，其中每个列表元素的形状为(144,2)元组，例如，一个元素包括[-102.97097952712484， 12.193768525539397]
/home/user/anaconda2/envs/testcaffe/lib/python2.7/site-packages/h5py/__init__.py:36: FutureWarning: Conversion of the second argument of issubdtype from `float` to `np.floating` is deprecated. In future, it will be treated as `np.float64 == np.dtype(float).type`.
from ._conv import register_converters as _register_converters
Traceback (most recent call last):
File "generalized-hough-demo.py", line 43, in <module>
acc = matchTable(im, table)
File "/home/user/workspace/jupyter_codes/PythonSIFT/Genarlized_Hough_Voting/generalised-hough-transform/match_table.py", line 38, in matchTable
acc[vector[0]+x, vector[1]+y]+=1
IndexError: only integers, slices (`:`), ellipsis (`...`), numpy.newaxis (`None`) and integer or boolean arrays are valid indices
我挣扎了两天，非常感谢你的专家意见。
答案 0 :(得分：-1)
此错误消息指出索引应为if(aTerm == null || aTerm.isZero()) {
return;// this is not a valid term
}
int insertIndex = -1;
int i = 0;
boolean addToList = true;
// loop until you've found your place, make sure to test for end conditions
// so you don't hit an IndexOutOfBounds error or create an infinite loop
while(insertIndex == -1 && i < theAL.size() && addToList) {
Term thisTerm = theAL.get(i);
if(aTerm.degree() == thisTerm.degree()) {
thisTerm.addIn(aTerm);
addToList = false;
} else if(aTerm.degree() > thisTerm.degree()) {
insertIndex = i;// this will end the loop, we know where to sort
// this element in our array
}
i++;// increment our counter
}
if(addToList) {
// make sure we found a valid location, if so, add in our term
if(insertIndex >= 0) {
theAL.add(insertIndex, aTerm);
} else {
theAL.add(aTerm); // insert the term at the end, it's less than all
// other terms in the list, or the list is empty
}
}
。如您所见，only integers是table的列表，每个vectors都是2d-vector，如您在问题中所述。
因此，vector和vector[0]主要是 float ，矩阵的索引必须是整数。这就是为什么它将错误指向 43 行。