Key Points Detection

Overview

Key Points Detection node locates user-defined keypoints and their connecting structure in the input frame using a custom-trained model.

Unlike Pose Estimation which targets human body joints, this node works with any keypoint schema defined by the loaded model — for example component pins on a PCB, fasteners on a product, or anatomical landmarks for non-human subjects. Use this node when Pose Estimation's fixed human skeleton does not fit your use case.

Input

Input Image

The image frame to analyze. Connect this to a camera or upstream image output.

Model Directory Path

Path to the keypoint model directory. See Model Directory Path for details.

Confidence Threshold

Minimum confidence score to keep a detected keypoint group. See Confidence Threshold for tuning guidance.

Overlap threshold

Boxes overlapping Filter (higher allows more overlap). See Overlap threshold for tuning guidance.

Overlay Results

Whether to draw keypoint markers and connecting lines on the output frame. See Overlay Results.

Draw Label Text

When enabled, draws each keypoint's name next to its marker on the overlay.

Custom Model Configs

Reveals extra framework and architecture fields for non-.nam models. See Custom Model Configs.

Model Color Format

Color format of the image data sent to the model. If left blank, the default for the chosen architecture is used. Adjust if your model was trained with a different color format.

• RGB - Red, Green, Blue channel order.
• BGR - Blue, Green, Red channel order.

Scale

Multiplicative factor for the R, G, and B channels of the image before sending to the neural network. If left blank, the default for the chosen architecture is used. Used to normalize pixel values to the range expected by the model.

Mean

Mean values for the R, G, and B channels that are subtracted from the image before sending to the neural network. If left blank, the default for the chosen architecture is used. Used to normalize the image to match the data distribution the model was trained on.

Input Format

Tensor format for the image data sent to the neural network. If left blank, the default for the chosen architecture is used.

• NHWC - Batch x Height x Width x Channel (channels last).
• NCHW - Batch x Channel x Height x Width (channels first).

Input Type

Data type of the image input sent to the neural network.

• Uint8 - Pixel values are integers in the range 0-255.
• Float32 - Pixel values are 32-bit floating point numbers, often in the range 0.0-1.0 or normalized according to the model's training.

Post Process Type

Method for processing the raw output from the neural network before producing keypoint results. If left blank, the default for the chosen architecture is used.

• keypoint_pose-movenet-single: MoveNet single-person format

"output_0": [1*1*17*3], [1_kp1_y, 1_kp1_x, 1_kp1_score, ..., 1_kp17_y, 1_kp17_x, 1_kp17_score]

• keypoint_pose-movenet-multiple: MoveNet multi-person format

"output_0": [1*N*56], [
 1_kp1_y, 1_kp1_x, 1_kp1_score, ..., 1_kp17_y, 1_kp17_x, 1_kp17_score,
 1_y1, 1_x1, 1_y2, 1_x2, 1_det_score,
 ...,
 n_kp1_y, n_kp1_x, n_kp1_score, ..., n_kp17_y, n_kp17_x, n_kp17_score,
 n_y1, n_x1, n_y2, n_x2, n_det_score
]

• keypoint_pose-ontf2: TensorFlow 2 keypoint detection format

"detection_classes": [N], [1_class_id, ..., n_class_id],
"detection_scores": [N], [1_score, ..., n_score],
"detection_boxes": [N*4], [1_y1, 1_x1, 1_y2, 1_x2, ..., n_y1, n_x1, n_y2, n_x2],
"detection_keypoints": [N*K*2], [1_kp1_y, 1_kp1_x, ..., 1_kpK_y, 1_kpK_x, ..., n_kp1_y, n_kp1_x, ..., n_kpK_y, n_kpK_x],
"detection_keypoints_scores": [N*K], [1_kp1_score, ..., 1_kpK_score, ..., n_kp1_score, ..., n_kpK_score]

• keypoint_efficientnet: Flat keypoint regressor format

"output_0": [K*2], [kp1_x, kp1_y, ..., kpK_x, kpK_y]

Output

Overlay Image

Output frame from the node. If overlays are enabled, keypoint markers, connecting lines, and optional labels are drawn on this frame.

Detected Count

Number of detected objects (keypoint groups) in the current frame.

Detected Objects

Array of keypoint detection objects. Each object contains:

• bbox array: Bounding box [x, y, width, height] around the detected instance.
• keypoints array: Array of keypoint objects, each with x, y, confidence, and optionally a name from the model schema.
• confidence number: Overall detection confidence.