我有一个txt文件,它有
x1,y1,z1,w1,desc1
x2,y2,z2,w2,desc2
till
xn,yn,zn,wn,descn
x、y、z和w是浮点数 desc是61个浮点数(不是数组或列表!图片中的图形仅用于说明)
我的代码是
for i in range(len(Xsnew)):
#Xsnew is all the x's in the txt file
x_features = Xsnew[i]
y_features = Ysnew[i]
z_features = Zsnew[i]
#print("x_features", x_features)
# Feature coordinates is the 3x1 matrix x1 y1 z1 in the picture
feature_coordinates = np.array([[x_features],[y_features],[z_features],[1]], dtype=float)
# P in the picture is the projection matrix
position_of_features = np.dot(projection_matrix, feature_coordinates)
u, v, w = position_of_features[0], position_of_features[1], position_of_features[2]
u_normalized = u / w
v_normalized = v / w
#print("U normalized is {} and V normalized is {}". format(u_normalized,v_normalized))
virtual_camera_desc = desc_features[i]
#print("virtual_camera_desc", virtual_camera_desc)
#print(" u = {} and v = {} and X_features = {} and desc = {}". format(u_normalized,v_normalized,x_features,virtual_camera_desc))
if 0 < u_normalized < image_height and 0 < v_normalized < image_width:
#print(len(desc), len(virtual_camera_desc))
#print("I SEE FEATURES")
bf = cv2.BFMatcher(cv2.NORM_HAMMING, crossCheck= True)
matches = bf.match(desc, virtual_camera_desc)
这很好,但速度太慢了。 我需要更快的算法,我在想怎么做“活”。但不确定,如何开始?有什么建议吗
谢谢:)
目前没有回答
相关问题 更多 >
编程相关推荐