class

TensorflowLite::Image::PoseEstimation::Output

Inherits TensorflowLite::Image::Detection::Points < TensorflowLite::Image::Detection::Lines < TensorflowLite::Image::Detection < JSON::Serializable < Reference < Object

Constants

LINES = [[BodyJoint::LeftEar, BodyJoint::LeftEye, BodyJoint::Nose, BodyJoint::RightEye, BodyJoint::RightEar], [BodyJoint::LeftWrist, BodyJoint::LeftElbow, BodyJoint::LeftShoulder, BodyJoint::RightShoulder, BodyJoint::RightElbow, BodyJoint::RightWrist], [BodyJoint::LeftShoulder, BodyJoint::LeftHip, BodyJoint::LeftKnee, BodyJoint::LeftAnkle], [BodyJoint::RightShoulder, BodyJoint::RightHip, BodyJoint::RightKnee, BodyJoint::RightAnkle], [BodyJoint::LeftHip, BodyJoint::RightHip]]

Constructors

new(points : Array(PoseEstimation::Point))

Source

Instance methods

lines

Source

markup(image : Canvas, minimum_score : Float32 = 0.3_f32) : Canvas

add the detection details to an image

if marking up the original image, you'll need to take into account how it was scaled and provide offsets

Source

points

Source

type

i.e. object, pose, face, age, class

Source