网站开发与运维收费明细,网站制作电话多少钱,网站建设是什么样的,什么网站能免费做推广1. LetterBox增强
当输入图片的尺寸和模型实际接收的尺寸可能不一致时#xff0c;通常需要使用LetterBox增强技术。具体步骤是先将图片按比例缩放#xff0c;将较长的边缩放到设定的尺寸以后#xff0c;再将较短的边进行填充#xff0c;最终短边的长度为stride的倍数即可。…1. LetterBox增强
当输入图片的尺寸和模型实际接收的尺寸可能不一致时通常需要使用LetterBox增强技术。具体步骤是先将图片按比例缩放将较长的边缩放到设定的尺寸以后再将较短的边进行填充最终短边的长度为stride的倍数即可。这种方法可以保留原始图像的纵横比同时还可以使图像更加适合目标检测算法的输入。 在YOLOv8代码中ultralytics/data/augment.py中class LetterBox类别实现了该功能。
import cv2
import numpy as npclass LetterBox:Resize image and padding for detection, instance segmentation, pose.def __init__(self, new_shape(640, 640), autoFalse, scaleFillFalse, scaleupTrue, centerTrue, stride32):Initialize LetterBox object with specific parameters.self.new_shape new_shapeself.auto autoself.scaleFill scaleFillself.scaleup scaleupself.stride strideself.center center # Put the image in the middle or top-leftdef __call__(self, labelsNone, imageNone):Return updated labels and image with added border.if labels is None:labels {}img labels.get(img) if image is None else imageshape img.shape[:2] # current shape [height, width]new_shape labels.pop(rect_shape, self.new_shape)if isinstance(new_shape, int):new_shape (new_shape, new_shape)# Scale ratio (new / old)r min(new_shape[0] / shape[0], new_shape[1] / shape[1])if not self.scaleup: # only scale down, do not scale up (for better val mAP)r min(r, 1.0)# Compute paddingratio r, r # width, height ratiosnew_unpad int(round(shape[1] * r)), int(round(shape[0] * r))dw, dh new_shape[1] - new_unpad[0], new_shape[0] - new_unpad[1] # wh paddingif self.auto: # minimum rectangledw, dh np.mod(dw, self.stride), np.mod(dh, self.stride) # wh paddingelif self.scaleFill: # stretchdw, dh 0.0, 0.0new_unpad (new_shape[1], new_shape[0])ratio new_shape[1] / shape[1], new_shape[0] / shape[0] # width, height ratiosif self.center:dw / 2 # divide padding into 2 sidesdh / 2if shape[::-1] ! new_unpad: # resizeimg cv2.resize(img, new_unpad, interpolationcv2.INTER_LINEAR)top, bottom int(round(dh - 0.1)) if self.center else 0, int(round(dh 0.1))left, right int(round(dw - 0.1)) if self.center else 0, int(round(dw 0.1))img cv2.copyMakeBorder(img, top, bottom, left, right, cv2.BORDER_CONSTANT, value(114, 114, 114)) # add borderif labels.get(ratio_pad):labels[ratio_pad] (labels[ratio_pad], (left, top)) # for evaluationif len(labels):labels self._update_labels(labels, ratio, dw, dh)labels[img] imglabels[resized_shape] new_shapereturn labelselse:return imgnew_shape(640, 640)
aug LetterBox(new_shape,centerTrue)labelsNone
img cv2.imread(./2.png,-1)lettered_img aug(labels,img)cv2.imshow(v8 letter_box,lettered_img)
cv2.waitKey(0)centerTrue
centerFalse
2 Mosaic增强
3 Mixup增强
