with open(output, 'wb') as outcsv:
writer = csv.writer(outcsv, dialect='excel')
header = [
'filename', 'distance', 'num_node', "score", "num_leaf",
"count_contourBoxes", "count_shareBox"
]
writer.writerow(header)
for row in data:
writer.writerow(row)
if __name__ == '__main__':
clfPath = "/Users/sephon/Desktop/Research/VizioMetrics/Corpus/Phylogenetic/CNN_corpus/line_patch/models/RF.pkl"
phyloParser = PhyloParser(clfPath=clfPath)
ground_truth_path = "/Users/sephon/Desktop/Research/VizioMetrics/Corpus/Phylogenetic/CNN_corpus/hq_ground_truth0228.csv"
folderPath = '/Users/sephon/Desktop/Research/VizioMetrics/Corpus/Phylogenetic/CNN_corpus/high_quality_tree'
ground_truth = {}
fileNameList = []
with open(ground_truth_path, 'rb') as incsv:
reader = csv.reader(incsv, dialect='excel')
reader.next()
for row in reader:
ground_truth[row[0]] = row[1]
# print row[0], row[1]
fileNameList.append(row[0])
# fileNameList = getFilesInFolder(folderPath)
outFileName = '/Users/sephon/Desktop/Research/VizioMetrics/Corpus/Phylogenetic/CNN_corpus/high_quality_tree_20170307_rline_new.csv'
viewPatch = getViewPatch(image, line)
# avgline = [int(avg_line_x), 0, int(avg_line_x), image.shape[0], 0]
# print avgline
# PhyloParser.displayLines(image, [avgline])
# PhyloParser.displayImage(viewPatch)
# PhyloParser.displayImage(hpatch)
# PhyloParser.displayImage(vpatch)
return feature, viewPatch
if __name__ == '__main__':
phyloParser = PhyloParser()
folderPath = '/Users/sephon/Desktop/Research/VizioMetrics/Corpus/Phylogenetic/CNN_corpus/far_species'
ground_truth = {}
fileNameList = []
fileList = []
for dirPath, dirNames, fileNames in os.walk(folderPath):
for f in fileNames:
extension = f.split('.')[-1].lower()
if extension in ["jpg", "png"]:
fileList.append(os.path.join(dirPath, f))
# fileNameList = getFilesInFolder(folderPath)
pathFolder = '/Users/sephon/Desktop/Research/VizioMetrics/Corpus/Phylogenetic/CNN_corpus/line_patch'
# datasetFlieName = '/Users/sephon/Desktop/Research/VizioMetrics/Corpus/Phylogenetic/CNN_corpus/line_feature.npy'
boundaries = [[335, 351, 185, 408], [351, 365, 212, 470]]
boxes = [[[336, 344, 382, 385], [335, 344, 315, 333], [336, 345, 337, 342],
[336, 345, 389, 408], [335, 345, 185, 257], [335, 349, 262, 310],
[335, 349, 346, 404]],
[[355, 362, 446, 448], [355, 363, 452, 470], [354, 363, 410, 435],
[355, 365, 212, 292], [354, 365, 298, 347], [354, 365, 346,
404]]]
boundaries, boxes = PhyloParser.stitchBoundries(boundaries, boxes)
print boundaries
print boxes
# new_boundaries, boxes = PhyloParser.splitBoundaries(boundaries, boxes)
phyloPaser = PhyloParser()
filename = '/Users/sephon/Desktop/Research/VizioMetrics/Corpus/Phylogenetic/phylogenetic_tree_for_parsing/PMC1474148_ijbsv02p0133g05.jpg'
filename = '/Users/sephon/Desktop/Research/VizioMetrics/Corpus/Phylogenetic/phylogenetic_tree_for_parsing/PMC2324105_1471-2148-8-100-1.jpg'
filename = '/Users/sephon/Desktop/Research/VizioMetrics/Corpus/Phylogenetic/phylogenetic_tree_for_parsing/PMC2323573_pone.0002033.g002.jpg'
filename = '/Users/sephon/Desktop/Research/VizioMetrics/Corpus/Phylogenetic/phylogenetic_tree_for_parsing/image_5569.jpg'
filename = '/Users/sephon/Desktop/Research/VizioMetrics/Corpus/Phylogenetic/phylogenetic_tree_for_parsing/PMC2323573_pone.0002033.g002.jpg'
# filename = '/Users/sephon/Desktop/Research/VizioMetrics/Corpus/Phylogenetic/phylogenetic_tree_for_parsing/PMC2761929_1471-2180-9-208-3.jpg'
# bg example
# filename = '/Users/sephon/Desktop/Research/VizioMetrics/Corpus/Phylogenetic/phylogenetic_tree_for_parsing/PMC2287175_1471-2148-8-57-2.jpg'
# filename = '/Users/sephon/Desktop/Research/VizioMetrics/Corpus/Phylogenetic/phylogenetic_tree_for_parsing/image_336.jpg'
# filename = '/Users/sephon/Desktop/Research/VizioMetrics/Corpus/Phylogenetic/phylogenetic_tree_for_parsing/image_337.jpg'
# filename = '/Users/sephon/Desktop/Research/VizioMetrics/Corpus/Phylogenetic/phylogenetic_tree_for_parsing/tree16.jpg'
def execute(filePath):
############ run experiment ######################
# a = ExperimentExecutor()
# a.autoRun('high_quality_tree/', 'groundTruth.csv', 'compareResult_0124.csv')
###################################################
# PMC2323573_pone.0002033.g002.jpg
try:
phyloParser = PhyloParser(clfPath=clfPath)
# folderPath = 'images/'
# folderPath = 'high_quality_tree/'
# fileNameList = getFilesInFolder(folderPath)
# fileNameList.sort()
# with open('image_fileName_list.txt', 'r')as f:
# fileNameList = f.readlines()
# testingFilePath = [join(folderPath, x) for x in fileNameList]
# result = {}
# groundTruth = {}
# a = ExperimentExecutor()
# a.execute(result, groundTruth, testingFilePath)
# for index, fileName in enumerate(fileNameList):
# print index, fileName
# # f.write(fileName + '\n')
# fileNameList = [x.rstrip() for x in fileNameList]
# fileNameList = [folderPath + x for x in fileNameList]
# for index in range(15, len(fileNameList)):
# try:
# print index, fileNameList[index]
# filePath = folderPath + fileNameList[index]
# fileName = fileNameList[index]
# # fileName = 'PMC2614190_ebo-04-181-g01.jpg'
# # fileName = 'PMC2674049_1471-2148-9-74-4.jpg'
# # fileName = 'PMC1326215_1471-2148-5-71-6.jpg'
# # fileName = 'PMC2644698_1471-2229-8-133-4.jpg'
# # fileName = 'PMC2644698_1471-2229-8-133-4.jpg'
# # fileName = 'PMC2775678_IJMB2009-701735.001.jpg'
# # fileName = 'PMC2697986_1471-2148-9-107-3.jpg'
# # fileName = 'PMC1182362_1471-2148-5-38-3.jpg'
# filePath = folderPath + fileName
# filePath = 'treeset/images/multi/1471-2148-10-52-2-l.jpg'
if isfile(filePath):
image = cv.imread(filePath, 0)
# PhyloParser.displayImage(image)
# imageData = ImageData(image)
# imageData = PhyloParser.preprocces(imageData, debug = False)
# # imageData = PhyloParser.detectLines__v2(imageData, debug=True)
# imageData = PhyloParser.testing(imageData)
image_data = ImageData(image)
image_data = phyloParser.preprocces(image_data, debug=False)
# PhyloParser.displayImage(image)
image_data = phyloParser.detectLines(image_data, debug=False)
# image_data = phyloPaser.traceTree(image_data, debug = True)
image_data = phyloParser.getCorners(image_data, debug=False)
image_data = phyloParser.makeLinesFromCorner(image_data, debug=False)
image_data = phyloParser.includeLinesFromCorners(image_data)
image_data = phyloParser.postProcessLines(image_data)
# image_data = phyloParser.removeRepeatedLines(image_data)
image_data = phyloParser.groupLines(image_data, debug=False)
image_data = phyloParser.matchLineGroups(image_data, debug=False)
# # image_data = phyloParser.detectCorners(image_data)
# # image_data = phyloParser.refineLinesByCorners(image_data)
# # image_data = phyloPaser.getSpecies_v2(image_data, debug = True)
# # image_data = phyloPaser.getSpecies(image_data, debug = True)
# image_data = phyloParser.matchLines(image_data, debug = True, useNew = False)
# image_data = phyloParser.getSpecies_v2(image_data, debug = True)
# # print image_data.orphanBox2Text
image_data = phyloParser.getSpecies_v3(image_data, debug=False)
# testString = phyloParser.makeTree(image_data, debug = True, tracing = False)
# treeString = phyloParser.constructTreeByTracing(image_data, debug = False)
treeString = phyloParser.constructTree_eval(image_data,
debug=False,
tracing=False)
dbdata = treeString.treeHead.getParentChildrenStyle()
# print dbdata['cluster']
with open('dbdata.csv', 'ab') as f:
csvwriter = csv.writer(f, delimiter='\t')
for i in range(dbdata['relation_index'] + 1):
if i >= dbdata['count_labels']:
string = ''
for node in dbdata[i]:
string += '#' + str(node)
csvwriter.writerow([fileName, str(i), string])
else:
csvwriter.writerow([fileName, str(i), dbdata[i]])
with open('dbdata_cluster.csv', 'ab') as f:
csvwriter = csv.writer(f, delimiter='\t')
if 'root' in dbdata['cluster']:
csvwriter.writerow([
fileName,
str(dbdata['relation_index']), dbdata['cluster']['root']
])
for i in range(dbdata['relation_index']):
csvwriter.writerow([fileName, str(i), dbdata['cluster'][i]])
except:
with open('error.txt', 'ab') as f:
f.write(filePath + '\n')
return fileNameList
def saveResult(data, output):
with open(output, 'wb') as outcsv:
writer = csv.writer(outcsv, dialect='excel')
header = ['filename', 'distance', 'num_node', "score"]
writer.writerow(header)
for row in data:
writer.writerow(row)
if __name__ == '__main__':
phyloParser = PhyloParser()
ground_truth_path = "/Users/sephon/Desktop/Research/VizioMetrics/Corpus/Phylogenetic/CNN_corpus/hq_ground_truth.csv"
ground_truth = {}
with open(ground_truth_path, 'rb') as incsv:
reader = csv.reader(incsv, dialect='excel')
reader.next()
for row in reader:
ground_truth[row[0]] = row[1]
# print row[0], row[1]
folderPath = '/Users/sephon/Desktop/Research/VizioMetrics/Corpus/Phylogenetic/CNN_corpus/high_quality_tree'
fileNameList = getFilesInFolder(folderPath)
outFileName = '/Users/sephon/Desktop/Research/VizioMetrics/Corpus/Phylogenetic/CNN_corpus/high_quality_tree_result_line_corners_0227.csv'
results = []
from ete3 import Tree
import operator
import random
from os import listdir
from os.path import isfile, join
from PhyloParser import *
def getFilesInFolder(folderPath):
fileNameList = [f for f in listdir(folderPath) if isfile(join(folderPath, f))]
return fileNameList
if __name__ == '__main__':
phyloPaser = PhyloParser()
folderPath = 'images/'
fileNameList = getFilesInFolder(folderPath)
for index, fileName in enumerate(fileNameList):
print index, fileName
for index in range(2, len(fileNameList)):
print index
filePath = folderPath + fileNameList[index]
if isfile(filePath) :
image = cv.imread(filePath,0)
image_data = ImageData(image)
image_data = phyloPaser.preprocces(image_data)