The new Mask Area Measurement block improves the experience when calculating the size of detected objects within a Workflow pipeline. For instance segmentation predictions, it computes the precise pixel area by counting non-zero mask pixels and properly handling holes. For standard object detection, it computes the area using bounding box dimensions (width × height).
By connecting the block to an object or segmentation model, you can apply a calibration value to convert pixel areas into real-world units (such as cm² or mm²). This block integrates with upstream perspective correction or camera calibration blocks, and downstream filtering blocks, enabling automated size-based filtering and quality control checks.
