New issue
Have a question about this project? Sign up for a free GitHub account to open an issue and contact its maintainers and the community.
By clicking “Sign up for GitHub”, you agree to our terms of service and privacy statement. We’ll occasionally send you account related emails.
Already on GitHub? Sign in to your account
运行get_features_into_CSV.py时报错 #3
Comments
Hi,
应该是图像文件读写格式有问题, 用你的 测试图片 单独跑下 get_128d_features() 看有没有问题
# 返回单张图像的128D特征
def get_128d_features(img_gray):
dets = detector(img_gray, 1)
if len(dets) != 0:
shape = predictor(img_gray, dets[0])
face_descriptor = facerec.compute_face_descriptor(img_gray, shape)
else:
face_descriptor = 0
return face_descriptor
get_128d_features("D:/test/test.jpg")
方便的话把你更改的源码贴给我下,应该不会有问题;
Best regrads
…------------------ 原始邮件 ------------------
发件人: "EstellaCheng"<notifications@github.com>;
发送时间: 2018年7月26日(星期四) 中午11:20
收件人: "coneypo/Dlib_face_recognition_from_camera"<Dlib_face_recognition_from_camera@noreply.github.com>;
抄送: "Subscribed"<subscribed@noreply.github.com>;
主题: [coneypo/Dlib_face_recognition_from_camera] 运行get_features_into_CSV.py时报错 (#3)
我clone代码到本地后,将get_from_camera里的照片清空,然后修改路径后,运行get_faces_from_camera后得到自己的数据集。然后把default_person.csv也删除了,再运行get_features_into_CSV.py,结果报错如下:
Traceback (most recent call last):
File "D:/360Downloads/JetBrains/Dlib_face_recognition_from_camera-master/get_features_into_CSV.py", line 76, in
write_into_csv(path_pics, path_csv + "default_person.csv")
File "D:/360Downloads/JetBrains/Dlib_face_recognition_from_camera-master/get_features_into_CSV.py", line 67, in write_into_csv
features_128d = return_128d_features(path_pics + dir_pics[i])
File "D:/360Downloads/JetBrains/Dlib_face_recognition_from_camera-master/get_features_into_CSV.py", line 43, in return_128d_features
face_descriptor = facerec.compute_face_descriptor(img_gray, shape)
RuntimeError: The full_object_detection must use the iBUG 300W 68 point face landmark style.
Process finished with exit code 1
我后来把图像换成博主原来的图像,再次运行还是报这个错。。。请博主赐教。
—
You are receiving this because you are subscribed to this thread.
Reply to this email directly, view it on GitHub, or mute the thread.
|
我单独运行了get_128d_features(),出现了问题。报错如下: 以下是我修改源码的地方: 保存line 28 path_save = "D:/360Downloads/JetBrains/Dlib_face_recognition_from_camera-master/data/get_from_camera/" 在get_features_into_CSV.py里面也是修改了路径,共三处: line 78 path_csv_rd = "D:/360Downloads/JetBrains/Dlib_face_recognition_from_camera-master/data/csvs/default_person.csv" 谢谢博主抽空帮我解决问题啦! |
Hi
不好意思刚下班
我觉得可能是你修改的时候误删了某一行,比如刚才发给你的那个 return_128d_features 函数,那样出错是因为传给他的图像格式不对,我觉得你可以check一下源码,类似于这种地方:
img=io.imread(path_img),看它传过去的是 文件路径 / img_gray 灰度图片 / RGB 通道图片 / HSV 通道图片 /..
from skimage import io
detector = dlib.get_frontal_face_detector()
predictor = dlib.shape_predictor('shape_predictor_68_face_landmarks.dat')
facerec = dlib.face_recognition_model_v1("dlib_face_recognition_resnet_model_v1.dat")
img_gray = io.imread(img_path)
img_gray = cv2.cvtColor(img_gray, cv2.COLOR_BGR2RGB)
dets = detector(img_gray, 1)
if len(dets) != 0:
shape = predictor(img_gray, dets[0])
face_descriptor = facerec.compute_face_descriptor(img_gray, shape)
else:
face_descriptor = 0
print(face_descriptor)
不行的话麻烦尝试下重新下载下 源码, link:
https://github.com/coneypo/Dlib_face_recognition_from_camera/blob/master/get_features_into_CSV.py
…------------------ 原始邮件 ------------------
发件人: "EstellaCheng"<notifications@github.com>;
发送时间: 2018年7月26日(星期四) 下午2:47
收件人: "coneypo/Dlib_face_recognition_from_camera"<Dlib_face_recognition_from_camera@noreply.github.com>;
抄送: "coneypo"<coneypo@gmail.com>; "Comment"<comment@noreply.github.com>;
主题: Re: [coneypo/Dlib_face_recognition_from_camera] 运行get_features_into_CSV.py时报错 (#3)
我单独运行了get_128d_features(),出现了问题。报错如下:
Traceback (most recent call last):
File "D:/360Downloads/JetBrains/Dlib_face_recognition_from_camera-master/test.py", line 24, in
get_128d_features("D:/360Downloads/JetBrains/Dlib_face_recognition_from_camera-master/data/get_from_camera/img_face_1.jpg")
File "D:/360Downloads/JetBrains/Dlib_face_recognition_from_camera-master/test.py", line 15, in get_128d_features
dets = detector(img_gray, 1)
RuntimeError: Unsupported image type, must be 8bit gray or RGB image.
以下是我修改源码的地方:
在get_faces_from_camera.py里面修改了存储路径。
保存
line 28 path_save = "D:/360Downloads/JetBrains/Dlib_face_recognition_from_camera-master/data/get_from_camera/"
在get_features_into_CSV.py里面也是修改了路径,共三处:
line 18 path_pics = "D:/360Downloads/JetBrains/Dlib_face_recognition_from_camera-master/data/get_from_camera/"
line 19 path_csv = "D:/360Downloads/JetBrains/Dlib_face_recognition_from_camera-master/data/csvs/"
line 78 path_csv_rd = "D:/360Downloads/JetBrains/Dlib_face_recognition_from_camera-master/data/csvs/default_person.csv"
谢谢博主抽空帮我解决问题啦!
—
You are receiving this because you commented.
Reply to this email directly, view it on GitHub, or mute the thread.
|
真是十分感谢博主啦。下了班还要帮我解决问题。。可是,,我把整个项目重新下载了一遍,就改了get_features_into_CSV.py里面的路径,运行起来还是会有哪个问题。。。。呜呜呜呜~~~天要灭我 |
你好,请 clone 最新的代码,这个 issue 应该不会复现了; |
我clone代码到本地后,将get_from_camera里的照片清空,然后修改路径后,运行get_faces_from_camera后得到自己的数据集。然后把default_person.csv也删除了,再运行get_features_into_CSV.py,结果报错如下:
Traceback (most recent call last):
File "D:/360Downloads/JetBrains/Dlib_face_recognition_from_camera-master/get_features_into_CSV.py", line 76, in
write_into_csv(path_pics, path_csv + "default_person.csv")
File "D:/360Downloads/JetBrains/Dlib_face_recognition_from_camera-master/get_features_into_CSV.py", line 67, in write_into_csv
features_128d = return_128d_features(path_pics + dir_pics[i])
File "D:/360Downloads/JetBrains/Dlib_face_recognition_from_camera-master/get_features_into_CSV.py", line 43, in return_128d_features
face_descriptor = facerec.compute_face_descriptor(img_gray, shape)
RuntimeError: The full_object_detection must use the iBUG 300W 68 point face landmark style.
Process finished with exit code 1
我后来把图像换成博主原来的图像,再次运行还是报这个错。。。请博主赐教。
The text was updated successfully, but these errors were encountered: