Visual Tools

This part of the library is designed for computer vision applications.

get_extensions

It calculates the count of each extension in given images.

Parameters	Datatype	Default Value
paths	list

Priority (in return)	Returns	Datatype	Condition
1	counts	dictionary	always

dataset_size_same

It checks that all images in the dataset have the same shape or not.

Parameters	Datatype	Default Value
paths	list

Priority (in return)	Returns	Datatype	Condition
1	same	boolean	always

dataset_ratio_same

It checks that all images in the dataset have the same ratio or not.

Parameters	Datatype	Default Value
paths	list

Priority (in return)	Returns	Datatype	Condition
1	same	boolean	always

crop

It crops the image from the centre in the requested shape.

Parameters	Datatype	Default Value
img	image object (OpenCV preferred)
path	string	None
crop_width	integer	256
crop_height	integer	256
get_img	boolean	False

Note

path holds the location info for the cropped image where it will be saved. If it is None, then the image is not being saved.

Priority (in return)	Returns	Datatype	Condition
1	img	OpenCV image object	get_img is True

fill

It fills with white the outside of the image to reach the desired shape.

Parameters	Datatype	Default Value
img	image object (OpenCV preferred)
path	string	None
fill_width	integer	256
fill_height	integer	256
get_img	boolean	False

Note

path holds the location info for the cropped image where it will be saved. If it is None, then the image is not being saved.

Priority (in return)	Returns	Datatype	Condition
1	img	OpenCV image object	get_img is True

make_square

It follows a predetermined workflow to fit all images inside the directory to the requested square shape.

Parameters	Datatype	Default Value
dir_from	string
dir_to	string
edge_len	integer	256

Attention

This function does not return anything.

dir_split

It splits images as train, test and split, then copies images to inner class folders. It is quite useful for image classification problems which are dealt by YOLO.

Parameters	Datatype	Default Value
src_dir	string
dest_dir	string
test_size	float
val_size	float

Note

directory at dest_dir might exist but it must be empty.

Attention

test_size and val_size must be less than 1, bigger than 0.

Attention

This function does not return anything.

cls_img_counter

It counts what class has how many images; this function is dataset organisation independent.

Parameters	Datatype	Default Value
dir_path	string

Priority (in return)	Returns	Datatype	Condition
1	class_amounts	dictionary	always

examine_sizes

It gives a brief about the sizes of the images in the dataset.

Parameters	Datatype	Default Value
paths	list

Attention

This function does not return anything.