Hi, saswat0! I want to ask you some questions. The program runs slowly on both my CPU and GPU. It takes four seconds to process an image. My cpu is 6 cores and 6 processes, and the memory is 8GB. My GPU is 8GB of TeslaM10. How many frames per second can you process, and what computing environment are you in

Hi:
Can you give a running video demo?
Best1

i tried to use the video.py on my local device , but shows me error

OpenCV Error: Unknown error code -10 (Raw image encoder error: Empty JPEG image (DNL not supported)) in throwOnEror, file /build/opencv-ys8xiq/opencv-2.4.9.1+dfsg/modules/highgui/src/grfmt_base.cpp, line 131 Traceback (most recent call last): File "video2.py", line 237, in <module> _, buffer = cv2.imencode('.jpg', reads) cv2.error: /build/opencv-ys8xiq/opencv-2.4.9.1+dfsg/modules/highgui/src/grfmt_base.cpp:131: error: (-10) Raw image encoder error: Empty JPEG image (DNL not supported) in function throwOnEror

i used the code of this repo https://github.com/sergiomsilva/alpr-unconstrained

and it worked for me but for the video.py in your repo it shows me that error , any idea to solve this problem?

Hi, saswat0. Thank you for your work, it's pretty! I want to ask you some problem. My environment is 8g GRID M10-8Q, with only one graphics card. If I use darknet and tensorflow compiled with gpu at the same time, the program will report an error:
Princess finished with exit code 137 (interrupted by signal 9: SIGKILL).
So my plate-detection is so slow, but vehicle detection and ocr are fast. Have you encountered such a problem, and how many frames can you process in one second?

Hi. I running your code on my comps, and it run well until we encounter one strange error: instead of numpy array, Image has string type. I try to convert it to numpy array with np.fromstring and it gets different error.

Is there any idea how to fix this? Thank you

Error:
Searching for vehicles using YOLO...
Scanning test_input/image.png
Traceback (most recent call last):
File "video.py", line 92, in
R,_ = detect(vehicle_net, vehicle_meta, img_path ,thresh=vehicle_threshold)
File "/home/lasbonjetson/Documents/License-Plate-Recognition/alpr/darknet/python/darknet.py", line 147, in detect
im, image = array_to_image(image)
File "/home/lasbonjetson/Documents/License-Plate-Recognition/alpr/darknet/python/darknet.py", line 18, in array_to_image
arr = arr.transpose(2, 0, 1)
AttributeError: 'str' object has no attribute 'transpose'.

Sorry, saswat0, I clicked in a wrong place that closed the issue. My cuda version is 10.2, so I install the cudatoolkit and cudnn with the command: conda install cudatoolkit=10.2, conda install cudnn. Because the first command specifies the cudatoolkit version, so the cudnn 7.6.5 is installed. Then I install the tensorflow-gpu with conda install tensorflow-gpu. The tensorflow-gpu 1.8 is installed. But when I run the python video.py, the gpu is not used.

I am getting errors related to opencv when i run "make".

Hi, saswat0! I have a 32g PH402 SKU 200 GPU on my friend's machine. It takes hundreds or even thousands of milliseconds for wpod-net to detect the license plate in a frame. It's too slow! How did you make the wpod-net run fast?

UnboundLocalError: local variable 'plate' referenced before assignment
this error comes up when testing with a different image.