Deploying nvOCDR to DeepStream

The nvOCDR library wraps the entire inference pipeline for optical character detection and recognition (OCR). This library consumes OCDNet and OCRNet models that are trained on TAO Toolkit. Whether you are building a surveillance system, a traffic monitoring application, or any other type of video-analytics solution, the nvOCDR library is an essential tool for achieving accurate and reliable results. This guide will walk through the steps for intgerating the nvOCDR library into DeepStream. Refer to the nvOCDR documentation for more information about nvOCDR.

Get Trained OCDNet and OCRNet Models

To deploy nvOCDR in DeepStream, you need to first train the OCDNet and OCRNet models with TAO Toolkit. You can either get started with models from the NVIDIA TAO Toolkit PTM (Pretrained Model) or can train your own model with TAO Toolkit. Refer to the training documentation for OCDNet and OCRNet to learn how to train your own model.

Download TAO Toolkit PTM from NGC

Note

Refer to NGC to set up your environment to run ngc commands.

You can download the pretrained OCDNet and OCRNet models with the following commands:

Copy
Copied!

            
            mkdir -p pretrained_models
ngc registry model download-version nvidia/tao/ocdnet:deployable_v1.0 --dest ./pretrained_models
ngc registry model download-version nvidia/tao/ocrnet:deployable_v1.0 --dest ./pretrained_models

A character_list.txt file is included with the pretrained OCRNet ONNX model. This is the vocabulary of the trained OCRNet model and is consumed by the nvOCDR library. Refer to the Preparing the Dataset section of the OCRNet documentation for more information about the character_list.txt.

Set up the DeepStream Development Environment

Once you have the trained models, you need to set up the DeepStream development environment.

On x86 platforms, you can start from the following container:

Copy
Copied!

            
            docker run --gpus=all -v <work_path>:<work_path> --rm -it --privileged --net=host nvcr.io/nvidia/deepstream:6.2-triton bash

# install opencv
apt update && apt install -y libopencv-dev

On Jetson platforms, you can start from the L4T container:

Copy
Copied!

            
            docker run --gpus=all -v <work_path>:<work_path> --rm -it --privileged --net=host nvcr.io/nvidia/deepstream-l4t:6.2-triton bash

# install opencv
apt update && apt install -y libopencv-dev

On Jetson platforms, you can also install Jetpack version 5.1 or greater and run the following command to install opencv:

Copy
Copied!

            
            # install opencv
apt update && apt install -y libopencv-dev

Next, you need to compile the TensorRT OSS plugin since OCDNet requires modulatedDeformConvPlugin:

Get the TensorRT repository:

Copy
Copied!

            
            git clone -b release/8.6 https://github.com/NVIDIA/TensorRT.git
cd TensorRT
git submodule update --init --recursive

Compile the TensorRT libnvinfer_plugin.so file:

Copy
Copied!

            
            mkdir build && cd build
# On X86 platform
cmake ..
# On Jetson platform
# cmake .. -DTRT_LIB_DIR=/usr/lib/aarch64-linux-gnu/
make nvinfer_plugin -j4

Copy the library to the system library path:

Copy
Copied!

            
            cp libnvinfer_plugin.so.8.6.0 /usr/lib/x86_64-linux-gnu/libnvinfer_plugin.so.8.5.2
# On Jetson platform:
# cp libnvinfer_plugin.so.8.6.0 /usr/lib/aarch64-linux-gnu/libnvinfer_plugin.so.8.5.2

Generate a TensorRT engine for OCDNet and OCRNet

After setting up the envrionment, you need to generate the TensorRT engine for the OCDNet and OCRNet models. This engine is used for running the models on the GPU. Use the following commands to generate TRT engines for OCDNet and OCRNet with dynamic batch size and specific height and weights:

Generate the OCDNet TRT engine with trtexec:

Copy
Copied!

            
            /usr/src/tensorrt/bin/trtexec --onnx=<path_to_pretrained ocdnet.onnx> --minShapes=input:1x3x736x1280 --optShapes=input:1x3x736x1280 --maxShapes=input:4x3x736x1280 --fp16 --saveEngine=<work_path>/ocdnet.fp16.engine

Generate the OCRNet TRT engine with trtexec:

Copy
Copied!

            
            /usr/src/tensorrt/bin/trtexec --onnx=<path_to_pretrained ocrnet.onnx> --minShapes=input:1x1x32x100 --optShapes=input:32x1x32x100 --maxShapes=input:32x1x32x100 --fp16 --saveEngine=<work_path>/ocrnet.fp16.engine

Build the nvOCDR Library and DeepStream Intermedia Library

Next, you need to build the nvOCDR library and the DeepStream intermedia library. These libraries are used for integrating the trained models into the DeepStream pipeline.

Get the nvOCDR repository:

Copy
Copied!

            
            git clone https://github.com/NVIDIA-AI-IOT/NVIDIA-Optical-Character-Detection-and-Recognition-Solution.git

Compile the libnvocdr.so nvOCDR library:

Copy
Copied!

            
            cd NVIDIA-Optical-Character-Detection-and-Recognition-Solution
make
export LD_LIBRARY_PATH=$(pwd)

Compile the libnvocdr_impl.so nvOCDR intermedia library for DeepStream:

Copy
Copied!

            
            cd deepstream
make

Run the nvOCDR DeepStream sample:

Finally, you can run the nvOCDR DeepStream sample to test the integration of the trained models into the DeepStream pipeline. You can build the DeepStream OCR pipeline with gst-launch-1.0 or run it with a C++ sample on github

Running the Pipeline with gst-launch-1.0

The following command runs a JPEG-image input pipeline with input batch-size=1. The output image will be saved to output.jpg:

Copy
Copied!

            
            gst-launch-1.0 filesrc location=<path_to_test_img> ! jpegparse ! nvv4l2decoder  ! \
m.sink_0 nvstreammux name=m batch-size=1 width=1280 height=1080 ! \
nvdsvideotemplate customlib-name=<path to libnvocdr_impl.so> \
customlib-props="ocdnet-engine-path:<path to ocdnet.fp16.engine>" \
customlib-props="ocdnet-input-shape:3,736,1280" \
customlib-props="ocdnet-binarize-threshold:0.1" \
customlib-props="ocdnet-polygon-threshold:0.3" \
customlib-props="ocdnet-max-candidate:200" \
customlib-props="ocrnet-engine-path:<path to ocrnet.fp16.engine>" \
customlib-props="ocrnet-dict-path:<path to character_list.txt>" \
customlib-props="ocrnet-input-shape:1,32,100" ! \
nvmultistreamtiler rows=1 columns=1 width=1280 height=720 ! nvvideoconvert ! nvdsosd ! \
nvvideoconvert ! 'video/x-raw,format=I420' ! jpegenc ! jpegparse ! filesink location=output.jpg

The following command runs a JPEG-image input pipeline with input batch-size=2:

Copy
Copied!

            
            gst-launch-1.0 filesrc location=<path_to_test_img> ! jpegparse ! nvv4l2decoder  ! \
m.sink_0 nvstreammux name=m batch-size=2 width=1280 height=1080 ! \
nvdsvideotemplate customlib-name=<path to libnvocdr_impl.so> \
customlib-props="ocdnet-engine-path:<path to ocdnet.fp16.engine>" \
customlib-props="ocdnet-input-shape:3,736,1280" \
customlib-props="ocdnet-binarize-threshold:0.1" \
customlib-props="ocdnet-polygon-threshold:0.3" \
customlib-props="ocdnet-max-candidate:200" \
customlib-props="ocrnet-engine-path:<path to ocrnet.fp16.engine>" \
customlib-props="ocrnet-dict-path:<path to character_list.txt>" \
customlib-props="ocrnet-input-shape:1,32,100" !  \
nvmultistreamtiler rows=1 columns=2 width=1280 height=720 ! nvvideoconvert ! nvdsosd ! \
nvvideoconvert ! 'video/x-raw,format=I420' ! jpegenc ! jpegparse ! filesink location=output.jpg \
filesrc location=<path to test image> ! jpegparse ! nvv4l2decoder  ! m.sink_1

The following command runs an MP4-video input pipeline with batch-size=1. The output video will be saved to output.mp4:

Copy
Copied!

            
            gst-launch-1.0 filesrc location=<path to test.mp4> ! qtdemux ! h264parse ! nvv4l2decoder ! \
m.sink_0 nvstreammux name=m batch-size=1 width=1280 height=1080 ! \
nvdsvideotemplate customlib-name=<path to libnvocdr_impl.so> \
customlib-props="ocdnet-engine-path:<path to ocdnet.fp16.engine>" \
customlib-props="ocdnet-input-shape:3,736,1280" \
customlib-props="ocdnet-binarize-threshold:0.1" \
customlib-props="ocdnet-polygon-threshold:0.3" \
customlib-props="ocdnet-max-candidate:200" \
customlib-props="ocrnet-engine-path:<path to ocrnet.fp16.engine>" \
customlib-props="ocrnet-dict-path:<path to character_list.txt>" \
customlib-props="ocrnet-input-shape:1,32,100" ! \
nvmultistreamtiler rows=1 columns=1 width=1280 height=720 ! nvvideoconvert ! nvdsosd ! \
nvvideoconvert ! 'video/x-raw(memory:NVMM),format=I420' ! nvv4l2h264enc ! h264parse ! \
mux.video_0 qtmux name=mux ! filesink location=output.mp4

You can download a test video from this link.

Configuring the nvOCDR library

You can configure the nvOCDR library parameters using the customlib-props arguments of nvdsvideotemplate. This is the template for setting the parameters:

Copy
Copied!

            
            nvdsvideotemplate customlib-name=libnvocdr_impl.so customlib-props="<nvOCDR attribute>:<nvOCDR attr value>"

Parameter	Data Type	Default	Description	Supported
`ocdnet-engine-path`	String	–	The absolute path to the OCDNet TensorRT engine	–
`ocdnet-input-shape`	String	–	The input shape (in CHW format) of the OCDNet TensorRT engine. Channel, height, and width are separated with commas.
`ocdnet-binarize-threshold`	Float	–	The threshold value to binarize the OCDNet output	>0
`ocdnet-unclip-ratio`	Float	1.5	The unclip ratio of the detected text region, which determines the output size	>0
`ocdnet-polygon-threshold`	Float	–	The threshold value to filter the polygons generated from the OCDNet postprocess based on the confidence score of polygons	[0, 1]
`ocdnet-max-candidate`	Unsigned int	–	The maximum output polygons from OCDNet	>0
`rectifier-upsidedown`	Unsigned int	0	The flag to enable upside-down processing in the Rectifier module. Set this option to 1 to enable nvOCDR to recognize characters totally upside down. The default value is 0.	0, 1
`ocrnet-engine-path`	String	–	The absolute path to the OCRNet TensorRT engine	–
`ocrnet-dict-path`	String	–	The absolute path to the OCRNet vocabulary file	–
`ocrnet-input-shape`	String	–	The input shape (in CHW format) of the OCRNet TensorRT engine. Channel, height, and width are separated with commas.

Integrating nvOCDR with Triton Inference Server

Triton Inference Server is an open source inference serving software that streamlines AI inferencing. Triton enables teams to deploy any AI model from multiple deep learning and machine learning frameworks, including TensorRT, TensorFlow, PyTorch, ONNX, OpenVINO, Python, RAPIDS FIL, and more.

This guide will walk through the steps for intgerating the nvOCDR library into Triton. Refer to the nvOCDR documentation for more information about nvOCDR.

Step1: Get the nvOCDR repository:

Copy
Copied!

            
            git clone https://github.com/NVIDIA-AI-IOT/NVIDIA-Optical-Character-Detection-and-Recognition-Solution.git

Step2: Download TAO Toolkit PTM from NGC

Note

Refer to NGC to set up your environment to run ngc commands.

You can download the pretrained OCDNet and OCRNet models with the following commands:
Copy

Copied!
```
            
            mkdir -p pretrained_models
ngc registry model download-version nvidia/tao/ocdnet:deployable_v1.0 --dest ./pretrained_models
ngc registry model download-version nvidia/tao/ocrnet:deployable_v1.0 --dest ./pretrained_models
        
```
A character_list.txt file is included with the pretrained OCRNet ONNX model. This is the vocabulary of the trained OCRNet model and is consumed by the nvOCDR library. Refer to the Preparing the Dataset section of the OCRNet documentation for more information about the character_list.txt.

Step3: Build triton server docker image:

Copy
Copied!

            
            cd NVIDIA-Optical-Character-Detection-and-Recognition-Solution/triton
bash setup_triton_server.sh [OCD input height] [OCD input width] [OCD input max batchsize] [DEVICE] [ocd onnx path> [ocr onnx path] [ocr character list path]

# For example
bash setup_triton_server.sh 736 1280 4 0 model/ocd.onnx model/ocr.onnx model/ocr_character_list

Step4: Build triton client docker image:

Copy
Copied!

            
            cd NVIDIA-Optical-Character-Detection-and-Recognition-Solution/triton
bash setup_triton_client.sh

Step5: Run nvocdr triton server:

Copy
Copied!

            
            docker run -it --net=host --gpus all --shm-size 8g nvcr.io/nvidian/tao/nvocdr_triton_server:v1.0 bash
CUDA_VISIBLE_DEVICES=<gpu idx> tritonserver --model-repository /opt/nvocdr/ocdr/triton/models/

Inference for high resolution images

nvocdr triton can support hight resolution images as input such as 4000x4000. you can change the spec file /opt/nvocdr/ocdr/triton/models/nvOCDR/spec.json in tirtion server container to support the high resolution images inference.
Copy

Copied!
```
            
            # to support high resolution images
is_high_resolution_input: true
        
```
Note

high resolution image inference only support batch size 1

Step6: Run nvocdr triton client:

open a new terminal and run commands below
Copy

Copied!
```
            
            docker run -it --rm -v <path to images dir>:<path to images dir>  --net=host nvcr.io/nvidian/tao/nvocdr_triton_client:v1.0 bash
python3 client.py -d <path to images dir> -bs 1
        
```
client.py args:
- -d : path to the images folder, the support image format includes ‘.jpg’, ‘.jpeg’, ‘.png’
- -bs: batch size for inference, only support batch size 1 when running the high resolution inference.

if you want to change some configures for triton server , you can exec the triton server container and stop the tritonserver process, then modify the spec file /opt/nvocdr/ocdr/triton/models/nvOCDR/spec.json. After you finished the modifing, you can launch the tritonserver and go to the client container to run the inference.

Below are the introduction of the parameters in spec file

Parameter	Data Type	Default	Description	Supported
`ocdnet_trt_engine_path`	String	–	The absolute path to the OCDNet TensorRT engine	–
`ocdnet_infer_input_shape`	List	–	The input shape (in CHW format) of the OCDNet TensorRT engine. Channel, height, and width are separated with commas.
`ocdnet-binarize-threshold`	Float	–	The threshold value to binarize the OCDNet output	>0
`ocdnet-unclip-ratio`	Float	1.5	The unclip ratio of the detected text region, which determines the output size	>0
`ocdnet-polygon-threshold`	Float	–	The threshold value to filter the polygons generated from the OCDNet postprocess based on the confidence score of polygons	[0, 1]
`ocdnet-max-candidate`	Unsigned int	–	The maximum output polygons from OCDNet	>0
`upsidedown`	bool	True	The flag to enable upside-down processing in the Rectifier module. Set this option to 1 to enable nvOCDR to recognize characters totally upside down. The default value is 0.	0, 1
`ocrnet_trt_engine_path`	String	–	The absolute path to the OCRNet TensorRT engine	–
`ocrnet_dict_file`	String	–	The absolute path to the OCRNet vocabulary file	–
`ocrnet_infer_input_shape`	List	–	The input shape (in CHW format) of the OCRNet TensorRT engine. Channel, height, and width are separated with commas.