def __init__(self):
# logging.debug("create PredictServer. loading model")
# rmac_model_proto = './features/deploy_resnet101_normpython.prototxt'
# rmac_model_weights = './features/rmac_model.caffemodel'
# print rmac_model_proto
# self._fDesc = RMACDescriptor(rmac_model_proto, rmac_model_weights)
self._fDesc = None
index_feat_path = os.path.join(VISUAL_FEATURE_DIR, 'finearts',
'rmac_index.csv')
self._searcher = Searcher(index_feat_path, cosine_distance)
logging.debug("model load successfully")
def _get_searcher(self, museum):
if museum not in self._searcher:
index_feat_path = os.path.join(MUSEUM_BASE_DIR,
MUSEUM_DATA_DIR[museum],
'visual_features', 'rmac_index.csv')
self._searcher[museum] = Searcher(index_feat_path, cosine_distance)
return self._searcher[museum]
class PredictServicer(rpc_pb2_grpc.PredictServicer):
def __init__(self):
# logging.debug("create PredictServer. loading model")
# rmac_model_proto = './features/deploy_resnet101_normpython.prototxt'
# rmac_model_weights = './features/rmac_model.caffemodel'
# print rmac_model_proto
# self._fDesc = RMACDescriptor(rmac_model_proto, rmac_model_weights)
self._fDesc = None
index_feat_path = os.path.join(VISUAL_FEATURE_DIR, 'finearts',
'rmac_index.csv')
self._searcher = Searcher(index_feat_path, cosine_distance)
logging.debug("model load successfully")
def getDescriptor(self):
if self._fDesc is None:
logging.debug("create PredictServer. loading model")
rmac_model_proto = './features/deploy_resnet101_normpython.prototxt'
rmac_model_weights = './features/rmac_model.caffemodel'
print rmac_model_proto
self._fDesc = RMACDescriptor(rmac_model_proto, rmac_model_weights)
return self._fDesc
def PredictPhoto(self, request, context):
#decode image from request
img_array = np.asarray(bytearray(request.data), dtype=np.uint8)
img = cv2.imdecode(img_array, cv2.IMREAD_COLOR)
t0 = time.time()
fdesc = self.getDescriptor()
features = fdesc.describe(img)
print('feature extraction took {:.3f}s').format(time.time() - t0)
t0 = time.time()
if features is not None:
results = self._searcher.search(features, 5)
print('Matching took {:.3f}s').format(time.time() - t0)
#print results
#process image
photo_info = "image shape:%s, size:%d, dtype:%s" % (
img.shape, img.size, img.dtype)
# print("predict photo request received. " + photo_info)
#write response
return rpc_pb2.PhotoPredictResponse(results=[
rpc_pb2.PhotoPredictResponse.Result(
text=photo_info,
image_url="%s/assets/%s" % (webhost, results[0][0]),
audio_url="/assets/audio/sample_0.4mb.mp3"),
rpc_pb2.PhotoPredictResponse.Result(text=photo_info,
image_url="%s/assets/%s" %
(webhost, results[1][0])),
rpc_pb2.PhotoPredictResponse.Result(text=photo_info,
image_url="%s/assets/%s" %
(webhost, results[2][0]))
])
ap.add_argument("-q", "--query", required=True, help="Path to the query image")
ap.add_argument("-r",
"--result_path",
required=True,
help="Path to the result path")
args = vars(ap.parse_args())
# initialize the image descriptor
cd = ColorDescriptor((8, 12, 3))
# load the query image and describe it
query = cv2.imread(args["query"])
features = cd.describe(query)
# perform the search
searcher = Searcher(args["index"])
results = searcher.search(features)
# display the query
fig = plt.figure()
ax = fig.add_subplot(4, 3, 1)
query = cv2.cvtColor(query, cv2.COLOR_BGR2RGB)
imgplot = plt.imshow(query)
i = 3
# loop over the results
for (score, resultID) in results:
i = i + 1
import cv2
import numpy as np
import time
import os
import _init_path
from features.rmacdescriptor import RMACDescriptor
from imagesearch.searcher import Searcher, cosine_distance
from features.feature_io import load_model
_ONE_DAY_IN_SECONDS = 60 * 60 * 24
VISUAL_FEATURE_DIR = '../../../../../../../work/data/visual_features'
rmac_model_proto = './features/deploy_resnet101_normpython.prototxt'
rmac_model_weights = './features/rmac_model.caffemodel'
print rmac_model_proto
fDesc = RMACDescriptor(rmac_model_proto, rmac_model_weights)
index_feat_path = os.path.join(VISUAL_FEATURE_DIR, 'finearts-annotated',
'rmac_index.csv')
searcher = Searcher(index_feat_path, cosine_distance)
img_file = '/home/pangolins/go/src/github.com/aitour/scene/web/assets/Images/museum/FineArtsQueryHighResolution/painting/587_small.JPG'
t0 = time.time()
features = fDesc.extract_feature(img_file)
print 'feature extraction took %.3f' % (time.time() - t0)
# USAGE
# python search.py --query query/100000.png --result-path image
from imagesearch.colordescriptor import ColorDescriptor
from imagesearch.searcher import Searcher
import argparse
import cv2
ap = argparse.ArgumentParser()
ap.add_argument("-q", "--query", required = True,
help = "path to the query image")
ap.add_argument("-r", "--result-path", required = True,
help = "Path to the result path")
args = vars(ap.parse_args())
cd = ColorDescriptor((8, 12, 3))
query = cv2.imread(args["query"])
features = cd.describe(query)
searcher = Searcher()
results = searcher.search(features)
cv2.imshow("Query", query)
for (score, resultID) in results:
result = cv2.imread(args["result_path"] + "/" + resultID)
cv2.imshow("Result", result)
cv2.waitKey(0)
"-d",
"--dataset",
required=True,
help="Path to the directory that contains the images we just indexed")
ap.add_argument("-i",
"--index",
required=True,
help="Path to where we stored our index")
args = vars(ap.parse_args())
index = args["index"]
# Load search index and perform image search.
with open(index, 'rb') as f:
pickle_file = f.read()
index = cPickle.loads(pickle_file)
searcher = Searcher(index)
# Test loop
for (query, queryFeatures) in index.items():
# Perform the search using the current query.
results = searcher.search(queryFeatures)
# Load the query image and display it.
path = args["dataset"] + "/%s" % (query)
queryImage = cv2.imread(path)
plt.imshow(queryImage)
plt.xticks([]), plt.yticks([]) # Hide the tick values on X and Y axis.
plt.show()
print("query: %s" % (query))
# Initialize the two montages to display the search results.
# Display the top 10 results with 5 images per output.
# USAGE
# python search.py --query query/100000.png --result-path image
from imagesearch.colordescriptor import ColorDescriptor
from imagesearch.searcher import Searcher
import argparse
import cv2
ap = argparse.ArgumentParser()
ap.add_argument("-q", "--query", required=True, help="path to the query image")
ap.add_argument("-r",
"--result-path",
required=True,
help="Path to the result path")
args = vars(ap.parse_args())
cd = ColorDescriptor((8, 12, 3))
query = cv2.imread(args["query"])
features = cd.describe(query)
searcher = Searcher()
results = searcher.search(features)
cv2.imshow("Query", query)
for (score, resultID) in results:
result = cv2.imread(args["result_path"] + "/" + resultID)
cv2.imshow("Result", result)
cv2.waitKey(0)