Welcome to Visionlib’s documentation!

Face Detection

class visionlib.face.detection.FDetector
This class contains all functions to detect face in an image.

…

Methods:

detect_face():

Used to detect face in an image. Returns the image with bounding boxes. Uses detector set by set_detector() method.

vdetect_face():

Used to detect face in a video. Yields the frame with bounding boxes. Uses detector set by set_detector() method.

set_detector():

Used to set detector to used by detect_face() method. If not set will use dnn detector as default.

detect_face(img=None, show=False, enable_gpu=False)

This method is used to detect face in an image.

Args:
img (numpy array)

This argument must the output which similar to opencv’s imread method’s output.

show (bool)

Set True to show image via cv2.imshow method.

Returns:
img (np.array)

Returns a numpy array of the image with bounding box.

box (list)

Returns x, y, w, h coordinates of the detected face Returns an empty list if no face is detected.

confidences (list)

Returns the associated confidences for the detected face.

set_detector(detector='dnn')

This method is used to set detector to be used to detect faces in an image.

Args:
detector (str)

The detector to be used. Can be any of the following: haar, hog, mtcnn, dnn. Dnn will be used as default.

vdetect_face(vid_path=None, show=False, enable_gpu=False, url=False)

This method is used to detect face in an video

Args:
vid_path (str)

Absolute Path to the video file.

show (bool)

Set True to show image via cv2.imshow method.

Yields:
img (np.array)

Returns a numpy array of the image with bounding box.

box (list) Returns x, y, w, h coordinates of the detected face Returns an empty list if no face is detected.

confidences (list)

Returns the associated confidences for the detected face.

Gender Detection

class visionlib.gender.detection.GDetector
This class contains all functions to detect gender of a given face

…

Methods:

detect_gender():

Used to detect gender from an face. Returns Predicted Gender and confidence.

detect_gender(img=None, enable_gpu=False)

This method is used to detect gender from an image.

Args:
img (numpy array)

This argument must the output which similar to opencv’s imread method’s output.

enable_gpu (bool) :

Set to True if You want to use gpu for prediction.

Returns:
str :

Returns the predicted gender.

int :

Returns the confidence for the predicted gender.

Object Detection

class visionlib.object.detection.detection.ODetection
This class contains all functions to detect objects from an image.

…

Methods:

detect_objects():

Used to detect objects from an image. Returns the bounding boxes, labels and confidence. Uses detector set by set_detector() method.

draw_box():

Used to draw the bounding box, labels and confidence in an image. Returns the frame with bounding boxes. Uses detector set by set_detector() method.

set_detector():

Used to set detector to used by detect_objects() method. If not set will use tiny yolo as default.

detect_objects(frame, enable_gpu=False)

This method is used to detect objects in an image.

Args:
frame (np.array):

Image to detect objects from.

enable_gpu (bool):

Set to true if You want to use gpu.

Returns:
list :

The detected bounding box.

list :

The detected class.

list :

Confidence for each detected class

draw_bbox(img, bbox, labels, confidence)

Draw’s Box around the detected objects.

Args
img (numpy.array):

The image to draw bounding boxes

bbox (list):

bounding boxes given detect_objects function.

labels (list):

labels given detect_objects function.

confidence (list):

Confidence for the detected label.

Returns
numpy.array :

The image with bounding boxes and labels.

set_detector(model_name='tiny_yolo', model_path=None, cfg_path=None, label_path=None)

Set’s the detector to use. Can be tiny-yolo or yolo. Setting to tiny-yolo will use yolov3-tiny. Setting to yolo will use yolov3.

Args:
model_name (str):

The model to use. If the given model is not present in pc, it will download and use it.

model_path (str):

Set this to path where the custom model You want to load is.

cfg_path (str):

Set this to path where the config file for custom model, You want to load is.

label_path (str):

Set this to path where the labels file for custom model, You want to load is.

Object Classification

class visionlib.object.classifier.inception_detector.Inception

This class is used to classify images using pre-trained Inception_v3 model.

Methods:

predict()

Used to predict the classes from the image.

predict(img=None, top=10)

This function is used to predict the possible classes in the image.

Args
img (numpy array)

The image for detecting classes

top (int)

The top n classes detected by the model to return

Returns
list

Returns a list of tuples with first value is the detected class and second value is the confidence in each tuple. eg: [(‘coffee_mug’, 88.80746364593506), (‘cup’, 9.322341531515121)]

Raises
AssertionError

Raised when no image is given

class visionlib.object.classifier.xception_detector.Xceptionv1

This class is used to classify images using pre-trained Xception_v1 model.

Methods:

predict()

Used to predict the classes from the image.

predict(img, top=10)

This function is used to predict the possible classes in the image.

Args
img (numpy array)

The image for detecting classes

top (int)

The top n classes detected by the model to return

Returns
list

Returns a list of tuples with first value is the detected class and second value is the confidence in each tuple. eg: [(‘coffee_mug’, 88.80746364593506), (‘cup’, 9.322341531515121)]

Raises
AssertionError

Raised when no image is given

class visionlib.object.classifier.vgg_detector.VGG

This class is used to classify images using pre-trained vgg16 model.

Methods:

predict()

Used to predict the classes from the image.

predict(img=None, top=10)

This function is used to predict the possible classes in the image.

Args
img (numpy array)

The image for detecting classes

top (int)

The top n classes detected by the model to return

Returns
list

Returns a list of tuples with first value is the detected class and second value is the confidence in each tuple. eg: [(‘coffee_mug’, 88.80746364593506), (‘cup’, 9.322341531515121)]

Raises
AssertionError

Raised when no image is given

class visionlib.object.classifier.detection.CDetector
set_detector(detector)

This method is used to set detector to be used to predict the classe in an image.

Args:
detector (str) :

The detector to be used. Can be any of the following: inception, xception, vgg. inception will be used as default.

Keypoint Detection

class visionlib.keypoints.detection.KDetector
This class contains functions for getting

keypoints for a space

Methods

set_detector()

Set the detector to detect keypoints

detect_keypoints()

Detect the keypoints in a face.

draw_points()

Draws the detected points in an image

detect_keypoints(img, rects=None, enable_gpu=False)

This function is used to detect keypoints.

Args
img (numpy array)

The image for detection.

rects (list)

The coordinates of face. (Not Needed for mtcnn detector)

enable_gpu (bool):

Set to True if you want to use gpu

Returns
list

Returns detected keypoint for each face

img

Returned only when no face is there for detection.

draw_points(img, points, show=False, color=0, 50, 255)
This function is used to draw the detected points

into the image.

Args
img (numpy array)

The image for drawing.

points (list)

The coordinates of the keypoints

show (bool)

Set to true if you want to display the image

color (tuple)

The color to display the points in.

Returns
numpy array

The image with keypoints marked

set_detector(detector)
This function is used to set the detector

to detect keypoints.

Args
detector (str)

Can be ‘dlib’ or ‘mtcnn’.

Raises
Invalid Selection

Raised when the selection is invalid

Indices and tables