License Plate Detection

Model by Open Source

This open source model detects vehicle license plates images and transcribes them using the TensorFlow Object Detection API. It takes as input license plate images in a PNG or JPG format. The model produces json file as output. This model can be used to identify vehicles for transportation planning and for traffic enforcement.
  • Description

    Product Description

    PERFORMANCE METRICS:

    53.6% Precision and 97.5% Recall

    The dataset used to train this model is not publicly available. This model uses precision and recall as its metrics. These are commonly used metrics in classification. This model achieves a recall of 97.52%, and precision of 53.64%. In practical terms this means that the model will most likely identify a license plate if it is present in an image but will also report about twice as many license plates as are truly present in a dataset.

    A higher precision score indicates that the majority of labels predicted by the model for different classes are accurate.

    Further information here.

    A higher recall score indicates that the model finds and predicts correct labels for the majority of the classes it is supposed to find.

    Further information here.

    OVERVIEW:

    The only training information available about this model at its source repository is that the model was based upon the ResNet architecture.

    TRAINING:

    Training dataset details are not publicly available.

    VALIDATION:

    Validation dataset details are not publicly available.

    INPUT SPECIFICATION

    The input(s) to this model must adhere to the following specifications:

    Filename Maximum Size Accepted Format(s)
    input.txt 1M .jpg, .png

    OUTPUT DETAILS

    This model will output the following:

    Filename Maximum Size Format
    results.json 1M .json