Using pete warden tutorials i had trained the inception network and training of which i am getting two files
1.retrained_graph.pb
2.retrained_label.txt
Using this i wanted to classify the flower image.
I had install pycharm and linked all the tensorflow library , i had also test the sample tensorflow code it is working fine.
Now when i run the label_image.py program which is
import tensorflow as tf, sys
image_path = sys.argv[1]
# Read in the image_data
image_data = tf.gfile.FastGFile(image_path, 'rb').read()
# Loads label file, strips off carriage return
label_lines = [line.rstrip() for line
in tf.gfile.GFile("/tf_files/retrained_labels.txt")]
# Unpersists graph from file
with tf.gfile.FastGFile("/tf_files/retrained_graph.pb", 'rb') as f:
graph_def = tf.GraphDef()
graph_def.ParseFromString(f.read())
_ = tf.import_graph_def(graph_def, name='')
with tf.Session() as sess:
# Feed the image_data as input to the graph and get first prediction
softmax_tensor = sess.graph.get_tensor_by_name('final_result:0')
predictions = sess.run(softmax_tensor, \
{'DecodeJpeg/contents:0': image_data})
# Sort to show labels of first prediction in order of confidence
top_k = predictions[0].argsort()[-len(predictions[0]):][::-1]
for node_id in top_k:
human_string = label_lines[node_id]
score = predictions[0][node_id]
print('%s (score = %.5f)' % (human_string, score))
i am getting this error message
/home/chandan/Tensorflow/bin/python /home/chandan/PycharmProjects/tf/tf_folder/tf_files/label_image.py
Traceback (most recent call last):
File "/home/chandan/PycharmProjects/tf/tf_folder/tf_files/label_image.py", line 7, in <module>
image_path = sys.argv[1]
IndexError: list index out of range
Could any one please help me with this issue.
You are getting this error because it is expecting image name (with path) as an argument.
In pycharm go to View->Tool windows->Terminal.
It is same as opening separate terminal. And run
python label_image.py /image_path/image_name.jpg
You are trying to get the command line argument by calling sys.argv[1]. So you need to give command line arguments to satisfy it. Looks like the argument required is a test image, you should pass its location as a parameter.
Pycharm should have a script parameters and interpreter options dialog which you can use to enter the required parameters.
Or you can call the script from a command line and enter the parameter via;
>python my_python_script.py my_python_parameter.jpg
EDIT:
According to the documents (I don't have pycharm installed on this computer), you should go to Run/Debug configuration menu and edit the configurations for your script. Add the absolute path of your file into Script Parameters box in quotes.
Or alternatively if you just want to skip the parameter thing completely, just get the path as raw_input (input in python3) or just simply give it to image_path = r"absolute_image_path.jpg"
Related
The following piece of code comes from Google's Vision API Documentation, the only modification I've made is adding the argument parser for the function at the bottom.
import argparse
import os
from google.cloud import vision
import io
def detect_text(path):
"""Detects text in the file."""
client = vision.ImageAnnotatorClient()
with io.open(path, 'rb') as image_file:
content = image_file.read()
image = vision.types.Image(content=content)
response = client.text_detection(image=image)
texts = response.text_annotations
print('Texts:')
for text in texts:
print('\n"{}"'.format(text.description))
vertices = (['({},{})'.format(vertex.x, vertex.y)
for vertex in text.bounding_poly.vertices])
print('bounds: {}'.format(','.join(vertices)))
ap = argparse.ArgumentParser()
ap.add_argument("-i", "--image", type=str,
help="path to input image")
args = vars(ap.parse_args())
detect_text(args)
If I run it from a terminal like below, I get this invalid file error:
PS C:\VisionTest> python visionTest.py --image C:\VisionTest\test.png
Traceback (most recent call last):
File "visionTest.py", line 31, in <module>
detect_text(args)
File "visionTest.py", line 10, in detect_text
with io.open(path, 'rb') as image_file:
TypeError: invalid file: {'image': 'C:\\VisionTest\\test.png'}
I've tried with various images and image types as well as running the code from different locations with no success.
Seems like either the file doesn't exist or is corrupt since it isn't even read. Can you try another image and validate it is in the location you expect?
I have a problem when using the ALDialog module on Python IDE and to load on Nao. I tried in different ways to load a dialogue but I always fall back on the same error.Runtimeerror LoadTopic::ALDialogIncorrect file myDialog.topIn the first case I write directly the text that I save in a. top file but at the time of LoadTopic () I have an error.In the second case I want to load the. top file by giving it the path. I come back to the same mistake again.Do you have a solution to my problem?Thank you very much.
import qi
import argparse
import os
import sys
from naoqi import ALProxy
def main(robot_ip, robot_port):
dialog = """
topic: ~myTopic() \n
language: enu \n
u:(test) hello \n """
file = open("myDialog.top","w")
file.write(dialog)
file.close()
# load topic
proxy = ALProxy("ALDialog",robot_ip,robot_port)
proxy.setLanguage("English")
self.topic = proxy.loadTopic("myDialog.top")
# start dialog
proxy.subscribe("myModule")
# activate dialog
proxy.activateTopic(self.topic)
if name == "main":
parser = argparse.ArgumentParser()
parser.add_argument("--ip", type=str,
default="169.254.245.164",help="Robot's IP address : '169.254.245.164'")
parser.add_argument("--port", type=int, default=9559,help="port number, the default value is OK in most cases")
args = parser.parse_args()
main(args.ip, args.port)
ALDialog.loadTopic expects an absolute filepath on the robot - it doesn't know anything about the context from which you're calling it (it could be from another computer, in which case of course it can't open that file). You need to be sure that your .top is indeed on the robot, and pass it's absolute path to ALDialog.
Once installed on the robot this path will be something like /home/nao/.local/share/PackageManager/apps/your-package-id/your-dialog-name/your-dialog-name_enu.top
I am new to GitPython and I am trying to get the content of a file within a commit. I am able to get each file from a specific commit, but I am getting an error each time I run the command. Now, I know that the file exist in GitPython, but each time I run my program, I am getting the following error:
returned non-zero exit status 1
I am using Python 2.7.6 and Ubuntu Linux 14.04.
I know that the file exist, since I also go directly into Git from the command line, check out the respective commit, search for the file, and find it. I also run the cat command on it, and the file contents are displayed. Many times when the error shows up, it says that the file in question does not exist. I am trying to go through each commit with GitPython, get every blob or file from each individual commit, and run an external Java program on the content of that file. The Java program is designed to return a string to Python. To capture the string returned from my Java code, I am also using subprocess.check_output. Any help will be greatly appreciated.
I tried passing in the command as a list:
cmd = ['java', '-classpath', '/home/rahkeemg/workspace/CSCI499_Java/bin/:/usr/local/lib/*:', 'java_gram.mainJava','absolute/path/to/file']
subprocess.check_output(cmd, stderr=subprocess.STDOUT, shell=False)
And I have also tried passing the command as a string:
subprocess.check_output('java -classpath /home/rahkeemg/workspace/CSCI499_Java/bin/:/usr/local/lib/*: java_gram.mainJava {file}'.format(file=entry.abspath.strip()), shell=True)
Is it possible to access the contents of a file from GitPython?
For example, say there is a commit and it has one file foo.java
In that file is the following lines of code:
foo.java
import java.io.FileInputStream;
import java.io.InputStream;
import java.util.ArrayList;
import java.util.List;
public class foo{
public static void main(String[] args) throws Exception{}
}
I want to access everything in the file and run an external program on it.
Any help would be greatly appreciated. Below is a piece of the code I am using to do so
#! usr/bin/env python
__author__ = 'rahkeemg'
from git import *
import git, json, subprocess, re
git_dir = '/home/rahkeemg/Documents/GitRepositories/WhereHows'
# make an instance of the repository from specified path
repo = Repo(path=git_dir)
heads = repo.heads # obtain the different repositories
master = heads.master # get the master repository
print master
# get all of the commits on the master branch
commits = list(repo.iter_commits(master))
cmd = ['java', '-classpath', '/home/rahkeemg/workspace/CSCI499_Java/bin/:/usr/local/lib/*:', 'java_gram.mainJava']
# start at the very 1st commit, or start at commit 0
for i in range(len(commits) - 1, 0, -1):
commit = commits[i]
commit_num = len(commits) - 1 - i
print commit_num, ": ", commit.hexsha, '\n', commit.message, '\n'
for entry in commit.tree.traverse():
if re.search(r'\.java', entry.path):
current_file = str(entry.abspath.strip())
# add the current file or blob to the list for the command to run
cmd.append(current_file)
print entry.abspath
try:
# This is the scenario where I pass arguments into command as a string
print subprocess.check_output('java -classpath /home/rahkeemg/workspace/CSCI499_Java/bin/:/usr/local/lib/*: java_gram.mainJava {file}'.format(file=entry.abspath.strip()), shell=True)
# scenario where I pass arguments into command as a list
j_response = subprocess.check_output(cmd, stderr=subprocess.STDOUT, shell=False)
except subprocess.CalledProcessError as e:
print "Error on file: ", current_file
# Use pop on list to remove the last string, which is the selected file at the moment, to make place for the next file.
cmd.pop()
First of all, when you traverse the commit history like this, the file will not be checked out. All you get is the filename, maybe leading to the file or maybe not, but certainly it will not lead to the file from different revision than currently checked-out.
However, there is a solution to this. Remember that in principle, anything you could do with some git command, you can do with GitPython.
To get file contents from specific revision, you can do the following, which I've taken from that page:
git show <treeish>:<file>
therefore, in GitPython:
file_contents = repo.git.show('{}:{}'.format(commit.hexsha, entry.path))
However, that still wouldn't make the file appear on disk. If you need some real path for the file, you can use tempfile:
f = tempfile.NamedTemporaryFile(delete=False)
f.write(file_contents)
f.close()
# at this point file with name f.name contains contents of
# the file from path entry.path at revision commit.hexsha
# your program launch goes here, use f.name as filename to be read
os.unlink(f.name) # delete the temp file
OK guys,
I'm currently working on a file reading and processing with Python & OpenCV cs' GUI feature. The feature will prompt the user to select a directory path for a folder containing 340 JPEG images, which I labelled them as "frame1" to "frame340". Then, I want to select several frames, process them, and save the processed ones in a different directory.
My big issue is, I'm trying to get only frame87, frame164, and frame248 from this folder with 340 images, and Python just keep returning error that claimed "directory name is invalid", like this:
Traceback (most recent call last):
File "C:\Users\maxwell_hamzah\Documents\Python27\imageReadBeta.py", line 25, in <module>
imgRead = os.listdir(str(dirname) + "/frame"+ str(i) + ".jpg")
WindowsError: [Error 267] The directory name is invalid: 'C:/Users/maxwell_hamzah/Documents/Python27/toby arm framed/frame87.jpg/*.*'
To help familiarize with the situation, here's what my work looks like:
import os
import numpy as np
import cv2
from matplotlib import pyplot as plt
from skimage import color, data, restoration
import Tkinter, tkFileDialog
# first, we setup the Tkinter features for file-reading
root = Tkinter.Tk()
root.withdraw()
# prompt user to ask about the file directory
dirname = tkFileDialog.askdirectory\
(parent=root,initialdir="/",title='Pick FRAMES directory')
X = [] # initiate an array to store read images
frameIndex = [87, 163, 248] #this index is which frames we are interested in
imgRead = ""
temp = []
# we begin to read only frame87, frame163, and frame248
for i in frameIndex:
imgRead = os.listdir(str(dirname) + "/frame"+ str(i) + ".jpg")
temp = cv2.imread(imgRead, -1)
X.append(temp)
I'm totally stuck on how to fix this bug on especially the for loop part, where the error comes from. Python keeps freeking out on the imgRead variable claiming that the directory is invalid. Plus, I'm also wondering on how to "export" processed files to other directories (e.g. saving processed images from "My Pictures" to "My Music")
Really appreciate your help, guys.
Maxwell
In the last block, you call a method to list files, which is expecting a directory, but you pass it a file path. That's a bug, and actually you don't need that here in the first place:
for i in frameIndex:
imgRead = "{0}/frame{1}.jpg".format(dirname, i)
temp = cv2.imread(imgRead, -1)
X.append(temp)
As to moving files in Python, that's a pretty classic need, there's plenty of doc out there. One example.
hi im trying to watermark a pdf fileusing pypdf2 though i get this error i cant figure out what goes wrong.
i get the following error:
Traceback (most recent call last): File "test.py", line 13, in <module>
page.mergePage(watermark.getPage(0)) File "C:\Python27\site-packages\PyPDF2\pdf.py", line 1594, in mergePage
self._mergePage(page2) File "C:\Python27\site-packages\PyPDF2\pdf.py", line 1651, in _mergePage
page2Content, rename, self.pdf) File "C:Python27\site-packages\PyPDF2\pdf.py", line 1547, in
_contentStreamRename
op = operands[i] KeyError: 0
using python 2.7.6 with pypdf2 1.19 on windows 32bit.
hopefully someone can tell me what i do wrong.
my python file:
from PyPDF2 import PdfFileWriter, PdfFileReader
output = PdfFileWriter()
input = PdfFileReader(open("test.pdf", "rb"))
watermark = PdfFileReader(open("watermark.pdf", "rb"))
# print how many pages input1 has:
print("test.pdf has %d pages." % input.getNumPages())
print("watermark.pdf has %d pages." % watermark.getNumPages())
# add page 0 from input, but first add a watermark from another PDF:
page = input.getPage(0)
page.mergePage(watermark.getPage(0))
output.addPage(page)
# finally, write "output" to document-output.pdf
outputStream = file("outputs.pdf", "wb")
output.write(outputStream)
outputStream.close()
Try writing to a StringIO object instead of a disk file. So, replace this:
outputStream = file("outputs.pdf", "wb")
output.write(outputStream)
outputStream.close()
with this:
outputStream = StringIO.StringIO()
output.write(outputStream) #write merged output to the StringIO object
outputStream.close()
If above code works, then you might be having file writing permission issues. For reference, look at the PyPDF working example in my article.
I encountered this error when attempting to use PyPDF2 to merge in a page which had been generated by reportlab, which used an inline image canvas.drawInlineImage(...), which stores the image in the object stream of the PDF. Other PDFs that use a similar technique for images might be affected in the same way -- effectively, the content stream of the PDF has a data object thrown into it where PyPDF2 doesn't expect it.
If you're able to, a solution can be to re-generate the source pdf, but to not use inline content-stream-stored images -- e.g. generate with canvas.drawImage(...) in reportlab.
Here's an issue about this on PyPDF2.