def check_neighbor(self, idx, cluster_idx, layer):
#if it is visited already
radius = self.search_radius
self.data_status_[idx] = False
result_set = RadiusNNResultSet(radius)
#print(np.shape(self.data_[idx]),(self.data_[idx]))
kdtree.kdtree_radius_search(self.root_, self.data_, result_set,
self.data_[idx])
nis = result_set.index_list
ndists = result_set.dist_list
#print("idx ",idx," nis",nis," ndists",ndists)
true_nis = np.count_nonzero(self.data_status_[nis])
if (true_nis > self.min_samples):
layer += 1
#print("set idx ",idx,"as label",cluster_idx)
self.data_label_[idx] = cluster_idx
for sub_idx in nis:
#print("nis" ,nis)
if self.data_status_[sub_idx] == True:
#print("set idx inner ",sub_idx,"as label",cluster_idx)
self.data_label_[sub_idx] = cluster_idx
self.check_neighbor(sub_idx, cluster_idx, layer)
#print("iter layer ",layer)
return layer
else:
#print("layer ",layer)
return layer
def KDTreeBenchmark(root_dir,
files,
k,
leaf_size,
radius,
feature=None,
feature2=None):
construction_time_sum = 0
knn_time_sum = 0
radius_time_sum = 0
brute_time_sum = 0
iteration_num = 0
for file in files:
if file.find('bin') == -1:
continue
iteration_num += 1
filename = os.path.join(root_dir, file)
db_np = read_velodyne_bin(filename)
begin_t = time.time()
root = kdtree.kdtree_construction(db_np, leaf_size, feature, feature2)
construction_time_sum += time.time() - begin_t
query = db_np[0, :]
begin_t = time.time()
result_set = KNNResultSet(capacity=k)
kdtree.kdtree_knn_search(root, db_np, result_set, query)
print("result set from KD Tree\n", result_set)
knn_time_sum += time.time() - begin_t
# print("--------")
begin_t = time.time()
result_set = RadiusNNResultSet(radius=radius)
kdtree.kdtree_radius_search(root, db_np, result_set, query)
#print(result_set)
radius_time_sum += time.time() - begin_t
#print("--------")
begin_t = time.time()
diff = np.linalg.norm(np.expand_dims(query, 0) - db_np, axis=1)
nn_idx = np.argsort(diff)
nn_dist = diff[nn_idx]
#print(nn_idx[0:k])
#print(nn_dist[0:k])
brute_time_sum += time.time() - begin_t
depth = [0]
max_depth = [0]
kdtree.traverse_kdtree(root, depth, max_depth)
print("tree depth: %d, max depth: %d" % (depth[0], max_depth[0]))
print("Kdtree: build %.3f, knn %.3f, radius %.3f, brute %.3f" %
(construction_time_sum * 1000 / iteration_num, knn_time_sum * 1000 /
iteration_num, radius_time_sum * 1000 / iteration_num,
brute_time_sum * 1000 / iteration_num))
def fit(self,data,method="radius"):
data_num = np.shape(data)[0]
print("number of data ",data_num)
self.simi_graph=np.zeros((data_num,data_num))
if method=="fully_connect":
for i in range(np.shape(data)[0]):
self.simi_graph[i] = get_dist_array(data,data[i])
#self.simi_graph=np.max(self.simi_graph)*2- self.simi_graph
self.simi_graph= gauss(self.simi_graph)
#for i in range(np.shape(data)[0]):
# self.simi_graph[i,i]=0
#print(self.simi_graph)
elif (method=="radius"):
root = kdtree.kdtree_construction(data, 16)
for di, datum in enumerate(data):
#print(datum)
result_set = RadiusNNResultSet(radius=10)
kdtree.kdtree_radius_search(root, data, result_set, datum)
nis = result_set.index_list
ndists = result_set.dist_list
#print("nis",nis)
#print("ndists",ndists)
for ni,ndist in zip(nis,ndists):
self.simi_graph[di][ni]=self.simi_graph[ni][di]=gauss(ndist,0.2)
#print("graph",self.simi_graph)
elif (method=="knn"):
print("knn is implemented")
tree=KDTree(data)
for di, datum in enumerate(data):
ndists,nis = tree.query([datum],20)
nis=nis[0]
ndists = ndists[0]
for ni,ndist in zip(nis,ndists):
if ni==di: continue
#print("HHHHHH",ni,di,ndist)
self.simi_graph[di][ni]=self.simi_graph[ni][di]=gauss(ndist)
else:
print("not available")
def main():
# configuration
leaf_size = 32
min_extent = 0.0001
k = 8
radius = 1
root_dir = './velodyne' # 数据集路径
cat = os.listdir(root_dir)
iteration_num = len(cat)
print("scipy-KDTree --------------")
construction_time_sum = 0
knn_time_sum = 0
radius_time_sum = 0
brute_time_sum = 0
for i in range(iteration_num):
filename = os.path.join(root_dir, cat[i])
db_np = read_velodyne_bin(filename)
# print(db_np.shape)
begin_t = time.time()
tree = KDTree(db_np, leafsize=leaf_size)
construction_time_sum += time.time() - begin_t
query = db_np[222, :]
begin_t = time.time()
distance, indices = tree.query(x=query, k=k)
print(distance)
print(indices)
knn_time_sum += time.time() - begin_t
begin_t = time.time()
diff = np.linalg.norm(np.expand_dims(query, 0) - db_np, axis=1)
nn_idx = np.argsort(diff)
nn_dist = diff[nn_idx]
brute_time_sum += time.time() - begin_t
# print(nn_idx)
print("Octree: build %.3f, knn %.3f, brute %.3f" % (construction_time_sum*1000/iteration_num,
knn_time_sum*1000/iteration_num,
brute_time_sum*1000/iteration_num))
print("octree --------------")
construction_time_sum = 0
knn_time_sum = 0
radius_time_sum = 0
brute_time_sum = 0
for i in range(iteration_num):
filename = os.path.join(root_dir, cat[i])
db_np = read_velodyne_bin(filename)
begin_t = time.time()
root = octree.octree_construction(db_np, leaf_size, min_extent)
construction_time_sum += time.time() - begin_t
query = db_np[222, :]
begin_t = time.time()
result_set = KNNResultSet(capacity=k)
octree.octree_knn_search(root, db_np, result_set, query)
print("result ------------")
print(result_set)
knn_time_sum += time.time() - begin_t
begin_t = time.time()
result_set = RadiusNNResultSet(radius=radius)
octree.octree_radius_search_fast(root, db_np, result_set, query)
# print("result ------------")
# print(result_set)
radius_time_sum += time.time() - begin_t
begin_t = time.time()
diff = np.linalg.norm(np.expand_dims(query, 0) - db_np, axis=1)
nn_idx = np.argsort(diff)
nn_dist = diff[nn_idx]
brute_time_sum += time.time() - begin_t
# print(nn_idx)
print("Octree: build %.3f, knn %.3f, radius %.3f, brute %.3f" % (construction_time_sum*1000/iteration_num,
knn_time_sum*1000/iteration_num,
radius_time_sum*1000/iteration_num,
brute_time_sum*1000/iteration_num))
otc_knn = knn_time_sum*1000
otc_radius = radius_time_sum*1000
otc_brute = brute_time_sum*1000
print("kdtree --------------")
construction_time_sum = 0
knn_time_sum = 0
radius_time_sum = 0
brute_time_sum = 0
for i in range(iteration_num):
filename = os.path.join(root_dir, cat[i])
db_np = read_velodyne_bin(filename)
begin_t = time.time()
root = kdtree.kdtree_construction(db_np, leaf_size)
construction_time_sum += time.time() - begin_t
query = db_np[222, :]
begin_t = time.time()
result_set = KNNResultSet(capacity=k)
kdtree.kdtree_knn_search(root, db_np, result_set, query)
knn_time_sum += time.time() - begin_t
print("result ------------")
print(result_set)
begin_t = time.time()
result_set = RadiusNNResultSet(radius=radius)
kdtree.kdtree_radius_search(root, db_np, result_set, query)
radius_time_sum += time.time() - begin_t
# print("result ------------")
# print(result_set)
begin_t = time.time()
diff = np.linalg.norm(np.expand_dims(query, 0) - db_np, axis=1)
nn_idx = np.argsort(diff)
nn_dist = diff[nn_idx]
brute_time_sum += time.time() - begin_t
# print(nn_idx)
kd_knn = knn_time_sum * 1000
kd_radius = radius_time_sum * 1000
kd_brute = brute_time_sum * 1000
print("Kdtree: build %.3f, knn %.3f, radius %.3f, brute %.3f" % (construction_time_sum * 1000 / iteration_num,
knn_time_sum * 1000 / iteration_num,
radius_time_sum * 1000 / iteration_num,
brute_time_sum * 1000 / iteration_num))
print("Octree knn/brute = %.3f, radius/bruet = %.3f" % (otc_knn/otc_brute, otc_radius/otc_brute))
print("Kdtree knn/brute = %.3f, radius/bruet = %.3f" % (kd_knn / kd_brute, kd_radius / kd_brute))
def get_neightbor_ids(self, root, query):
result_set = RadiusNNResultSet(self.radius)
kdtree.kdtree_radius_search(root, self.data_, result_set, query)
nis = result_set.index_list
return nis
def main():
# configuration
db_size = 64000
dim = 3
leaf_size = 32
min_extent = 0.0001
k = 8
radius = 1
iteration_num = 100
print("octree --------------")
construction_time_sum = 0
knn_time_sum = 0
radius_time_sum = 0
brute_time_sum = 0
for i in range(iteration_num):
db_np = (np.random.rand(db_size, dim) - 0.5) * 100
begin_t = time.time()
root = octree.octree_construction(db_np, leaf_size, min_extent)
construction_time_sum += time.time() - begin_t
query = np.random.rand(3)
begin_t = time.time()
result_set = KNNResultSet(capacity=k)
octree.octree_knn_search(root, db_np, result_set, query)
knn_time_sum += time.time() - begin_t
begin_t = time.time()
result_set = RadiusNNResultSet(radius=radius)
octree.octree_radius_search_fast(root, db_np, result_set, query)
radius_time_sum += time.time() - begin_t
begin_t = time.time()
diff = np.linalg.norm(np.expand_dims(query, 0) - db_np, axis=1)
nn_idx = np.argsort(diff)
nn_dist = diff[nn_idx]
brute_time_sum += time.time() - begin_t
print("Octree: build %.3f, knn %.3f, radius %.3f, brute %.3f" %
(construction_time_sum * 1000 / iteration_num, knn_time_sum * 1000 /
iteration_num, radius_time_sum * 1000 / iteration_num,
brute_time_sum * 1000 / iteration_num))
print("kdtree --------------")
construction_time_sum = 0
knn_time_sum = 0
radius_time_sum = 0
brute_time_sum = 0
for i in range(iteration_num):
db_np = np.random.rand(db_size, dim)
begin_t = time.time()
root = kdtree.kdtree_construction(db_np, leaf_size)
construction_time_sum += time.time() - begin_t
query = np.random.rand(3)
begin_t = time.time()
result_set = KNNResultSet(capacity=k)
kdtree.kdtree_knn_search(root, db_np, result_set, query)
knn_time_sum += time.time() - begin_t
begin_t = time.time()
result_set = RadiusNNResultSet(radius=radius)
kdtree.kdtree_radius_search(root, db_np, result_set, query)
radius_time_sum += time.time() - begin_t
begin_t = time.time()
diff = np.linalg.norm(np.expand_dims(query, 0) - db_np, axis=1)
nn_idx = np.argsort(diff)
nn_dist = diff[nn_idx]
brute_time_sum += time.time() - begin_t
print("Kdtree: build %.3f, knn %.3f, radius %.3f, brute %.3f" %
(construction_time_sum * 1000 / iteration_num, knn_time_sum * 1000 /
iteration_num, radius_time_sum * 1000 / iteration_num,
brute_time_sum * 1000 / iteration_num))
def main():
# configuration
leaf_size = 32
min_extent = 0.0001
k = 8
radius = 1
# data
root_dir = '../data/bin'
cat = os.listdir(root_dir)
iteration_num = len(cat)
print("Brute-force --------------")
brute_time_sum = 0
for i in range(iteration_num):
filename = os.path.join(root_dir, cat[i])
db_np = read_velodyne_bin(filename)
query = db_np[0, :]
begin_t = time.time()
diff = np.linalg.norm(np.expand_dims(query, 0) - db_np, axis=1)
nn_idx = np.argsort(diff)
nn_dist = diff[nn_idx]
brute_time_sum += time.time() - begin_t
print("Brute-force: %.3f" % (brute_time_sum * 1000 / iteration_num))
# scipy.spatial.KDTree
print("scipy.spatial.KDTree --------------")
construction_time_sum = 0
knn_time_sum = 0
for i in range(iteration_num):
filename = os.path.join(root_dir, cat[i])
db_np = read_velodyne_bin(filename)
begin_t = time.time()
tree = spatial.KDTree(db_np, leaf_size)
construction_time_sum += time.time() - begin_t
query = db_np[0, :]
begin_t = time.time()
res = tree.query(query, k)
knn_time_sum += time.time() - begin_t
print("scipy.spatial.KDTree: build %.3f, knn %.3f" %
(construction_time_sum * 1000 / iteration_num,
knn_time_sum * 1000 / iteration_num))
print("Octree --------------")
construction_time_sum = 0
knn_time_sum = 0
radius_time_sum = 0
brute_time_sum = 0
for i in range(iteration_num):
filename = os.path.join(root_dir, cat[i])
db_np = read_velodyne_bin(filename)
begin_t = time.time()
root = octree.octree_construction(db_np, leaf_size, min_extent)
construction_time_sum += time.time() - begin_t
query = db_np[0, :]
begin_t = time.time()
result_set = KNNResultSet(capacity=k)
octree.octree_knn_search(root, db_np, result_set, query)
knn_time_sum += time.time() - begin_t
begin_t = time.time()
result_set = RadiusNNResultSet(radius=radius)
octree.octree_radius_search_fast(root, db_np, result_set, query)
radius_time_sum += time.time() - begin_t
begin_t = time.time()
diff = np.linalg.norm(np.expand_dims(query, 0) - db_np, axis=1)
nn_idx = np.argsort(diff)
nn_dist = diff[nn_idx]
brute_time_sum += time.time() - begin_t
print("Octree: build %.3f, knn %.3f, radius %.3f, brute %.3f" %
(construction_time_sum * 1000 / iteration_num, knn_time_sum * 1000 /
iteration_num, radius_time_sum * 1000 / iteration_num,
brute_time_sum * 1000 / iteration_num))
print("KD-Tree --------------")
construction_time_sum = 0
knn_time_sum = 0
radius_time_sum = 0
brute_time_sum = 0
for i in range(iteration_num):
filename = os.path.join(root_dir, cat[i])
db_np = read_velodyne_bin(filename)
begin_t = time.time()
root = kdtree.kdtree_construction(db_np, leaf_size)
construction_time_sum += time.time() - begin_t
query = db_np[0, :]
begin_t = time.time()
result_set = KNNResultSet(capacity=k)
kdtree.kdtree_knn_search(root, db_np, result_set, query)
knn_time_sum += time.time() - begin_t
begin_t = time.time()
result_set = RadiusNNResultSet(radius=radius)
kdtree.kdtree_radius_search(root, db_np, result_set, query)
radius_time_sum += time.time() - begin_t
begin_t = time.time()
diff = np.linalg.norm(np.expand_dims(query, 0) - db_np, axis=1)
nn_idx = np.argsort(diff)
nn_dist = diff[nn_idx]
brute_time_sum += time.time() - begin_t
print("KD-Tree: build %.3f, knn %.3f, radius %.3f, brute %.3f" %
(construction_time_sum * 1000 / iteration_num, knn_time_sum * 1000 /
iteration_num, radius_time_sum * 1000 / iteration_num,
brute_time_sum * 1000 / iteration_num))
def main():
# configuration
leaf_size = 32
min_extent = 0.001
k = 8
radius = 1
root_dir = '/home/xcy/myWork/pointCloudProcessing/lesson2/bin' # 数据集路径
cat = os.listdir(root_dir)
iteration_num = len(cat)
print("octree --------------")
construction_time_sum = 0
knn_time_sum = 0
radius_time_sum = 0
radius_fast_time_sum = 0
brute_time_sum = 0
for i in range(iteration_num):
filename = os.path.join(root_dir, cat[i])
db_np = read_velodyne_bin(filename)
begin_t = time.time()
root = octree.octree_construction(db_np, leaf_size, min_extent)
construction_time_sum += time.time() - begin_t
query = db_np[0, :]
begin_t = time.time()
result_set = KNNResultSet(capacity=k)
octree.octree_knn_search(root, db_np, result_set, query)
knn_time_sum += time.time() - begin_t
#
begin_t = time.time()
result_set = RadiusNNResultSet(radius=radius)
octree.octree_radius_search(root, db_np, result_set, query)
radius_time_sum += time.time() - begin_t
begin_t = time.time()
result_set = RadiusNNResultSet(radius=radius)
octree.octree_radius_search_fast(root, db_np, result_set, query)
radius_fast_time_sum += time.time() - begin_t
begin_t = time.time()
diff = np.linalg.norm(np.expand_dims(query, 0) - db_np, axis=1)
nn_idx = np.argsort(diff)
nn_dist = diff[nn_idx]
brute_time_sum += time.time() - begin_t
print(
"Octree: build %.3f, knn %.3f, radius %.3f, radius_fast %.3f, brute %.3f"
% (construction_time_sum * 1000 / iteration_num, knn_time_sum * 1000 /
iteration_num, radius_time_sum * 1000 / iteration_num,
radius_fast_time_sum * 1000 / iteration_num,
brute_time_sum * 1000 / iteration_num))
print("kdtree --------------")
construction_time_sum = 0
knn_time_sum = 0
radius_time_sum = 0
brute_time_sum = 0
for i in range(iteration_num):
filename = os.path.join(root_dir, cat[i])
db_np = read_velodyne_bin(filename)
begin_t = time.time()
root = kdtree.kdtree_construction(db_np, leaf_size)
construction_time_sum += time.time() - begin_t
# begin_t = time.time()
# root = kdtree.kdtree_construction_median(db_np, leaf_size)
# construction_time_sum += time.time() - begin_t
query = db_np[0, :]
begin_t = time.time()
diff = np.linalg.norm(np.expand_dims(query, 0) - db_np, axis=1)
nn_idx = np.argsort(diff)
nn_dist = diff[nn_idx]
brute_time_sum += time.time() - begin_t
begin_t = time.time()
result_set = KNNResultSet(capacity=k)
kdtree.kdtree_knn_search(root, db_np, result_set, query)
knn_time_sum += time.time() - begin_t
begin_t = time.time()
result_set = RadiusNNResultSet(radius=radius)
kdtree.kdtree_radius_search(root, db_np, result_set, query)
radius_time_sum += time.time() - begin_t
print("Kdtree: build %.3f, knn %.3f, radius %.3f, brute %.3f" %
(construction_time_sum * 1000 / iteration_num, knn_time_sum * 1000 /
iteration_num / (brute_time_sum * 1000 / iteration_num),
radius_time_sum * 1000 / iteration_num /
(brute_time_sum * 1000 / iteration_num),
brute_time_sum * 1000 / iteration_num))
def main():
# configuration
leaf_size = 32
min_extent = 0.0001
k = 8
radius = 1
filename = '000000.bin'
db_np = read_velodyne_bin(filename)
iteration_num = 1
cat = []
print("octree --------------")
construction_time_sum = 0
knn_time_sum = 0
radius_time_sum = 0
brute_time_sum = 0
for i in range(iteration_num):
db_np = read_velodyne_bin(filename)
begin_t = time.time()
root = octree.octree_construction(db_np, leaf_size, min_extent)
construction_time_sum += time.time() - begin_t
query = db_np[0, :]
begin_t = time.time()
result_set = KNNResultSet(capacity=k)
octree.octree_knn_search(root, db_np, result_set, query)
knn_time_sum += time.time() - begin_t
begin_t = time.time()
result_set = RadiusNNResultSet(radius=radius)
octree.octree_radius_search_fast(root, db_np, result_set, query)
radius_time_sum += time.time() - begin_t
begin_t = time.time()
diff = np.linalg.norm(np.expand_dims(query, 0) - db_np, axis=1)
nn_idx = np.argsort(diff)
nn_dist = diff[nn_idx]
brute_time_sum += time.time() - begin_t
print("Octree: build %.3f, knn %.3f, radius %.3f, brute %.3f" %
(construction_time_sum * 1000 / iteration_num, knn_time_sum * 1000 /
iteration_num, radius_time_sum * 1000 / iteration_num,
brute_time_sum * 1000 / iteration_num))
print("kdtree --------------")
construction_time_sum = 0
knn_time_sum = 0
radius_time_sum = 0
brute_time_sum = 0
for i in range(iteration_num):
#filename = os.path.join(root_dir, cat[i])
db_np = read_velodyne_bin(filename)
begin_t = time.time()
root = kdtree.kdtree_construction(db_np, leaf_size)
construction_time_sum += time.time() - begin_t
query = db_np[0, :]
begin_t = time.time()
result_set = KNNResultSet(capacity=k)
kdtree.kdtree_knn_search(root, db_np, result_set, query)
knn_time_sum += time.time() - begin_t
begin_t = time.time()
result_set = RadiusNNResultSet(radius=radius)
kdtree.kdtree_radius_search(root, db_np, result_set, query)
radius_time_sum += time.time() - begin_t
begin_t = time.time()
diff = np.linalg.norm(np.expand_dims(query, 0) - db_np, axis=1)
nn_idx = np.argsort(diff)
nn_dist = diff[nn_idx]
brute_time_sum += time.time() - begin_t
print("Kdtree: build %.3f, knn %.3f, radius %.3f, brute %.3f" %
(construction_time_sum * 1000 / iteration_num, knn_time_sum * 1000 /
iteration_num, radius_time_sum * 1000 / iteration_num,
brute_time_sum * 1000 / iteration_num))
def main():
# configuration
leaf_size = 32
min_extent = 0.005
k = 8
radius = 1
nums = 1 # 点的数量（搜索）
abs_path = os.path.abspath(
os.path.dirname(__file__)) # 数据集路径 (数据集直接放在当前路径下)
filename = os.path.join(abs_path, '000000.bin') # 只读取一个文件,如果要读取所有文件,需要循环读入
iteration_num = 1
# 读取数据并进行采样
db_np_origin = read_velodyne_bin(filename) # N*3
db_np_idx = np.random.choice(db_np_origin.shape[0],
size=(30000, )) # 随机采样30000个点
db_np = db_np_origin[db_np_idx]
#root_dir = '/Users/renqian/cloud_lesson/kitti' # 数据集路径
#cat = os.listdir(root_dir)
#iteration_num = len(cat)
print("搜索数据集中的 %d 个点: " % nums)
print(
"---------------------------------Octree ----------------------------------"
)
construction_time_sum = 0
knn_time_sum = 0
radius_time_sum = 0
brute_time_sum = 0
for i in range(iteration_num):
#filename = os.path.join(root_dir, cat[i])
'''
db_np_origin = read_velodyne_bin(filename) # N*3
db_np_idx = np.random.choice(db_np_origin.shape[0],size =(30000,)) # 随机采样30000个点
db_np = db_np_origin[db_np_idx]
'''
begin_t = time.time()
root = octree.octree_construction(db_np, leaf_size,
min_extent) # Octree Construction
construction_time_sum += time.time() - begin_t
print("The sum of the points: ", db_np.shape[0])
nums = db_np.shape[0]
# 查找nums个点
for num in range(nums):
query = db_np[num, :] # 搜索点数组（假设第一个点为搜索点）
# knn search
begin_t = time.time()
result_set = KNNResultSet(capacity=k)
octree.octree_knn_search(root, db_np, result_set, query)
knn_time_sum += time.time() - begin_t
# radius search
begin_t = time.time()
result_set = RadiusNNResultSet(radius=radius)
#octree.octree_radius_search_fast(root, db_np, result_set, query)
octree.octree_radius_search(root,
db_np,
result_set,
query,
search='fast')
radius_time_sum += time.time() - begin_t
# brute search
begin_t = time.time()
diff = np.linalg.norm(np.expand_dims(query, 0) - db_np, axis=1)
nn_idx = np.argsort(diff)
nn_dist = diff[nn_idx]
brute_time_sum += time.time() - begin_t
print("Octree: build %.3f, knn %.3f, radius %.3f, brute %.3f" %
(construction_time_sum * 1000 / iteration_num, knn_time_sum *
1000 / iteration_num, radius_time_sum * 1000 / iteration_num,
brute_time_sum * 1000 / iteration_num))
print(
"---------------------------------Kdtree ----------------------------------"
)
construction_time_sum = 0
knn_time_sum = 0
radius_time_sum = 0
brute_time_sum = 0
knn_scipy_time_sum = 0
radius_scipy_time_sum = 0
for i in range(iteration_num):
#filename = os.path.join(root_dir, cat[i])
#db_np = read_velodyne_bin(filename) # N*3
print("The sum of the points: ", db_np.shape[0])
begin_t = time.time()
root = kdtree.kdtree_construction(db_np,
leaf_size) # kd tree construction
construction_time_sum += time.time() - begin_t
tree = spatial.KDTree(db_np) # 利用spatial库函数构造kdtree
for num in range(nums):
query = db_np[num, :]
# scipy.spatial.KDtree
begin_t = time.time()
tree.query(query, k=k)
knn_scipy_time_sum += time.time() - begin_t
#print(tree.query(query, k=k))
# knn search
begin_t = time.time()
result_set = KNNResultSet(capacity=k)
kdtree.kdtree_knn_search(root, db_np, result_set, query)
knn_time_sum += time.time() - begin_t
#print(result_set)
# scipy.spatial.KDtree (radius search)
begin_t = time.time()
tree.query_ball_point(query, radius)
radius_scipy_time_sum += time.time() - begin_t
# radius search
begin_t = time.time()
result_set = RadiusNNResultSet(radius=radius)
kdtree.kdtree_radius_search(root, db_np, result_set, query)
radius_time_sum += time.time() - begin_t
# print(result_set)
# brute search
begin_t = time.time()
diff = np.linalg.norm(np.expand_dims(query, 0) - db_np, axis=1)
nn_idx = np.argsort(diff)
nn_dist = diff[nn_idx]
brute_time_sum += time.time() - begin_t
print(
"Kdtree: build %.3f, knn_scipy %.3f, knn %.3f, radius_scipy %.3f, radius %.3f, brute %.3f"
%
(construction_time_sum * 1000 / iteration_num, knn_scipy_time_sum *
1000 / iteration_num, knn_time_sum * 1000 / iteration_num,
radius_scipy_time_sum * 1000 / iteration_num, radius_time_sum *
1000 / iteration_num, brute_time_sum * 1000 / iteration_num))
def main():
# configuration
leaf_size = 32
min_extent = 0.0001
k = 8
radius = 1
root_dir = '/home/gc/github_repository/DeepBlue/lesson_2/kitty/' # 数据集路径
cat = os.listdir(root_dir)
iteration_num = len(cat)
print("octree --------------")
construction_time_sum = 0
knn_time_sum = 0
radius_time_sum = 0
brute_time_sum = 0
for i in range(iteration_num):
filename = os.path.join(root_dir, cat[i])
db_np = read_velodyne_bin(filename)
db_np = db_np.T
begin_t = time.time()
root = octree.octree_construction(db_np, leaf_size, min_extent)
construction_time_sum += time.time() - begin_t
query = db_np[0,:]
begin_t = time.time()
result_set = KNNResultSet(capacity=k)
octree.octree_knn_search(root, db_np, result_set, query)
knn_time_sum += time.time() - begin_t
begin_t = time.time()
result_set = RadiusNNResultSet(radius=radius)
octree.octree_radius_search_fast(root, db_np, result_set, query)
radius_time_sum += time.time() - begin_t
begin_t = time.time()
diff = np.linalg.norm(np.expand_dims(query, 0) - db_np, axis=1)
nn_idx = np.argsort(diff)
nn_dist = diff[nn_idx]
brute_time_sum += time.time() - begin_t
print("Octree: build %.3f, knn %.3f, radius %.3f, brute %.3f" % (construction_time_sum*1000/iteration_num,
knn_time_sum*1000/iteration_num,
radius_time_sum*1000/iteration_num,
brute_time_sum*1000/iteration_num))
otc_knn = knn_time_sum*1000
otc_radius = radius_time_sum*1000
otc_brute = brute_time_sum*1000
print("kdtree --------------")
construction_time_sum = 0
knn_time_sum = 0
radius_time_sum = 0
brute_time_sum = 0
for i in range(iteration_num):
filename = os.path.join(root_dir, cat[i])
db_np = read_velodyne_bin(filename)
begin_t = time.time()
root = kdtree.kdtree_construction(db_np, leaf_size)
construction_time_sum += time.time() - begin_t
query = db_np[0,:]
begin_t = time.time()
result_set = KNNResultSet(capacity=k)
kdtree.kdtree_knn_search(root, db_np, result_set, query)
knn_time_sum += time.time() - begin_t
begin_t = time.time()
result_set = RadiusNNResultSet(radius=radius)
kdtree.kdtree_radius_search(root, db_np, result_set, query)
radius_time_sum += time.time() - begin_t
begin_t = time.time()
diff = np.linalg.norm(np.expand_dims(query, 0) - db_np, axis=1)
nn_idx = np.argsort(diff)
nn_dist = diff[nn_idx]
brute_time_sum += time.time() - begin_t
print("Kdtree: build %.3f, knn %.3f, radius %.3f, brute %.3f" % (construction_time_sum * 1000 / iteration_num,
knn_time_sum * 1000 / iteration_num,
radius_time_sum * 1000 / iteration_num,
brute_time_sum * 1000 / iteration_num))
kd_knn = knn_time_sum * 1000
kd_radius = radius_time_sum * 1000
kd_brute = brute_time_sum * 1000
print("Octree knn/brute = %.3f, radius/bruet = %.3f" % (otc_knn/otc_brute, otc_radius/otc_brute) )
print("Kdtree knn/brute = %.3f, radius/bruet = %.3f" % (kd_knn / otc_brute, kd_radius / otc_brute) )
def main():
# configuration
leaf_size = 32
min_extent = 0.0001
k = 8
radius = 1
#root_dir = '/Users/renqian/cloud_lesson/kitti' # 数据集路径
#cat = os.listdir(root_dir)
#iteration_num = len(cat)
# load date
filename = '/home/ljn/SLAM/dateset/000000.bin'
db_np = read_velodyne_bin(filename)
iteration = 1
print("octree -----------------------------------")
construction_time_sum = 0
knn_time_sum = 0
radius_time_sum = 0
brute_time_sum = 0
result_set_knn = KNNResultSet(capacity=k)
query = db_np[95, :]
begin_t = time.time()
root = octree.octree_construction(db_np, leaf_size, min_extent)
construction_time_sum += time.time() - begin_t
begin_t = time.time()
octree.octree_knn_search(root, db_np, result_set_knn, query)
knn_time_sum += time.time() - begin_t
print('knn search result\n', result_set_knn)
begin_t = time.time()
result_set_rnn = RadiusNNResultSet(radius=radius)
octree.octree_radius_search_fast(root, db_np, result_set_rnn, query)
radius_time_sum += time.time() - begin_t
print('rnn search result\n', result_set_rnn)
begin_t = time.time()
diff = np.linalg.norm(np.expand_dims(query, 0) - db_np, axis=1)
nn_idx = np.argsort(diff)
nn_dist = diff[nn_idx]
nn_dist_idx_pre = np.linspace(0, nn_dist.shape[0], nn_dist.shape[0])
nn_dist_idx = nn_dist_idx_pre[nn_idx]
brute_time_sum += time.time() - begin_t
#brute knn search
result_set_knn_brute = KNNResultSet(capacity=k)
for index in range(k):
result_set_knn_brute.add_point(nn_dist[index], nn_dist_idx[index])
# brute radiusNN search
result_set_rnn_brute = RadiusNNResultSet(radius=radius)
for index in range(nn_dist.shape[0]):
if nn_dist[index] < radius:
result_set_rnn_brute.add_point(nn_dist[index], nn_dist_idx[index])
continue
else:
break
print("Octree: build %.3f, knn %.3f, radius %.3f, brute %.3f" %
(construction_time_sum * 1000, knn_time_sum * 1000,
radius_time_sum * 1000, brute_time_sum * 1000))
print("kdtree -----------------------------------")
construction_time_sum = 0
knn_time_sum = 0
radius_time_sum = 0
brute_time_sum = 0
for i in range(iteration):
query = db_np[95, :]
begin_t = time.time()
root = kdtree.kdtree_construction(db_np, leaf_size)
construction_time_sum += time.time() - begin_t
begin_t = time.time()
result_set_knn = KNNResultSet(capacity=k)
kdtree.kdtree_knn_search(root, db_np, result_set_knn, query)
knn_time_sum += time.time() - begin_t
begin_t = time.time()
result_set_rnn = RadiusNNResultSet(radius=radius)
kdtree.kdtree_radius_search(root, db_np, result_set_rnn, query)
radius_time_sum += time.time() - begin_t
begin_t = time.time()
diff = np.linalg.norm(np.expand_dims(query, 0) - db_np, axis=1)
nn_idx = np.argsort(diff)
nn_dist = diff[nn_idx]
brute_time_sum += time.time() - begin_t
nn_dist_idx_pre = np.linspace(0, nn_dist.shape[0] - 1,
nn_dist.shape[0])
nn_dist_idx = nn_dist_idx_pre[nn_idx]
# brute knn search
result_set_knn_brute = KNNResultSet(capacity=k)
for index in range(k):
result_set_knn_brute.add_point(nn_dist[index], nn_dist_idx[index])
# brute radiusNN search
result_set_rnn_brute = RadiusNNResultSet(radius=radius)
for index in range(nn_dist.shape[0]):
if nn_dist[index] < radius:
result_set_rnn_brute.add_point(nn_dist[index],
nn_dist_idx[index])
continue
else:
break
print("Kdtree: build %.3f, knn %.3f, radius %.3f, brute %.3f" %
(construction_time_sum * 1000, knn_time_sum * 1000,
radius_time_sum * 1000, brute_time_sum * 1000))
print("scipy kdtree -----------------------------------")
construction_time_sum = 0
knn_time_sum = 0
radius_time_sum = 0
query = db_np[95, :]
begin_t = time.time()
tree = spatial.KDTree(db_np, leaf_size)
construction_time_sum += time.time() - begin_t
#no knn
begin_t = time.time()
knn_time_sum += time.time() - begin_t
begin_t = time.time()
result_set_rnn = tree.query_ball_point(query, radius)
radius_time_sum += time.time() - begin_t
print('rnn search result\n', result_set_rnn)
print("Octree: build %.3f, knn %.3f, radius %.3f" %
(construction_time_sum * 1000, knn_time_sum * 1000,
radius_time_sum * 1000))