RectDetectorPipeline
Esta página aún no está traducida. Se muestra la versión en inglés.
RectDetectorPipeline is an interface for your own vision processors that find rectangles in the camera image. It extends the FTC VisionProcessor, so it works with any VisionPortal. It lives in the package dev.easyfamily.EasyStack.controllable.
#Methods
getDetections(): the list of rectangles found in the last frame. You must implement it.suspendDetection(): stops looking for rectangles. It does nothing by default.resumeDetection(): starts looking again. It does nothing by default.
#RectBlob
Each detection is a RectBlob. All sizes are normalized, so they are values between 0 and 1 that do not depend on the camera resolution.
- id: the number of the detection.
- label: the name, or Rectangle when none is given.
- cxNorm, cyNorm: the center of the rectangle, from 0 on the left and top to 1 on the right and bottom.
- wNorm, hNorm: the width and height.
- confidence: how sure the pipeline is.
area(): width times height, never negative.
Java
for (RectDetectorPipeline.RectBlob blob : pipeline.getDetections()) {
if (blob.confidence > 0.7 && blob.area() > 0.02) {
double turn = (blob.cxNorm - 0.5) * 0.8;
drive.driveRobotCentric(0, 0, turn);
}
}Última actualización 2026-11-16