API

car_speed.car_speed_detection.check_images_match(read_dir, image_count)

Check if the images in the directory ‘read_dir’ match the image count ‘image_count’.

Parameters

  • read_path (str) – Path to the directory that stored the images.

  • image_count (int) – Image count users expected.

Returns

True if matches, False otherwise.

Return type

bool

car_speed.car_speed_detection.combine_video_and_speed(video_path, speed_path, output_path='combined_video_and_speed.mp4')

Read in the speed of the car from ‘speed_path’, print the corresponding spped on each frame in the video at ‘video_path’, and output the video at ‘output_path’.

Parameters

  • video_path (str) – Path to the video.

  • speed_path (str) – Path to the speed text.

  • output_path (str) – Path to the output.

Returns

True if combining success, False if combining fail.

Return type

bool

car_speed.car_speed_detection.preprocess(read_dir, train_path, output_path, resize=0.5, x_slice=8, y_slice=6)

Preprocess all the images stored in directory ‘read_dir’ and attach each frame with the corresponding speed stored at ‘train_path’ and output the feature set to a txt file at ‘output_path’.

Parameters

  • read_dir (str) – Path to the directory that stored the images.

  • train_path (str) – Path to the file that store the speed of each frame in the video.

  • output_path (str) – Path to the output file that stored the feature set.

  • resize (int) – Ratio factor along horizontal and vertical axis.

  • x_slice (int) – Desired slice along horizontal axis.

  • y_slice (int) – Desired slice along vertical axis.

Returns

Time it takes to preprocess the images.

Return type

float

car_speed.car_speed_detection.read(read_path, write_dir)

Load a mp4 video from ‘read_path’ and stored each frame as a jpg file in a new directory ‘write_dir’.

Parameters

  • read_path (str) – Path to the input video, should be a mp4 file.

  • param (str) – Path to the output directory.

Returns

The index of the last frames being processed.

Return type

int

car_speed.car_speed_detection.speed_detection(model_path, video, output_path, required_resize, required_x_slice, required_y_slice, MEAN_CONST, STD_CONST)

Detect the speed of the automobile using the pretrained model from ‘model_path’ and input video from ‘video’, and then output the txt file of the detected speed at ‘output_path’.

Parameters

  • model_path (str) – Path to the pretrained model.

  • video (str) – Path to the video.

  • output_path (str) – Path to the output.

  • required_resize (int) – Resize scale that was used for the pretrained model.

  • required_x_slice – x slice that was used for the pretrained model.

  • required_y_slice – y slice that was used for the pretrained model.

  • MEAN_CONST – Mean of the training set, used to normalize the testing set.

  • STD_CONST – Standard Deviation of the training set, used to normalize the testing set.

Returns

tuple containing:

(index (int): index of the last frame that is predicted, detection_time (float): Total time took to predict the speed of the car from the input video).

Return type

tuple

car_speed.car_speed_detection.train(read_path, validation_split=0.75, batch_size=128, epoch=100, verbose=1)

Train the Artifitial Neural Network (ANN) model with the preprocessed feature set stored in ‘read_path’ and stored the model as ‘Model.h5’ in the root directory.

Parameters

  • read_path (str) – Path to the file that stores the feature set.

  • validation_split (float) – Float between 0 and 1. Fraction of the training data to be used as validation data. The model will set apart this fraction of the training data, will not train on it, and will evaluate the loss and any model metrics on this data at the end of each epoch.

  • batch_size (str) – Integer or None. Number of samples per gradient update.

  • epoch (int) – Integer. Number of epochs to train the model. An epoch is an iteration over the entire x and y data provided. Note that in conjunction with initial_epoch, epochs is to be understood as “final epoch”. The model is not trained for a number of iterations given by epochs, but merely until the epoch of index epochs is reached.

  • verbose (int) – 0, 1, or 2. Verbosity mode. 0 = silent, 1 = progress bar, 2 = one line per epoch. Note that the progress bar is not particularly useful when logged to a file, so verbose=2 is recommended when not running interactively.

Returns

tuple containing:

(mse (float): mean square error of the model that is tested with validation data, MEAN_CONST (float): mean value that is used to normalize the feature set, STD_CONST (float): standard deviation value that is used to normalize the feature set).

Return type

tuple