psenet-text-detector 0.1.1

Fast and accurate text detection library built on PSENet implementation

PSENet: Shape Robust Text Detection with Progressive Scale Expansion Network

Packaged Version of the Pytorch implementation of PSENet text detector

Overview

PSENet is designed as a segmentation-based detector with multiple predictions for each text instance. These predictions correspond to different `kernels' produced by shrinking the original text instance into various scales. Consequently, the final detection can be conducted through our progressive scale expansion algorithm which gradually expands the kernels with minimal scales to the text instances with maximal and complete shapes.

Getting started

Installation

• Install using conda for Linux, Mac and Windows (preferred):

conda install -c fcakyon psenet-text-detector

• Install using pip for Linux and Mac:

pip install psenet-text-detector

Basic Usage

# import package
import psenet_text_detector as psenet

# set image path and export folder directory
image_path = 'figures/idcard.png'
output_dir = 'outputs/'

# apply craft text detection and export detected regions to output directory
prediction_result = psenet.detect_text(image_path, output_dir, cuda=False)

Advanced Usage

# import package
import psenet_text_detector as psenet

# set image path and export folder directory
image_path = 'figures/idcard.png'
output_dir = 'outputs/'

# read image
image = psenet.read_image(image_path)

# load model
psenet_model = psenet.load_psenet_model()

# perform prediction
prediction_result = psenet.get_prediction(image=image,
                               		  model=psenet_model,
                                       	  binary_th=1.0,
                                       	  kernel_num=3,
                                       	  upsample_scale=1,
                                       	  long_size=1280,
                                       	  min_kernel_area=10.0,
                                       	  min_area=300.0,
                                       	  min_score=0.93,
                                       	  cuda=True)

# export detected text regions
exported_file_paths = psenet.export_detected_regions(image_path,
                                              	    image,
                                              	    boxes=prediction_result["boxes"],
                                              	    output_dir=output_dir)

# export box visualization
_ = psenet.visualize_detection(image_path,
            		       image=image,
        		       quads=prediction_result["boxes"],
                    	       output_dir=output_dir)