NKL

NKL（NanKai Lines 的缩写）是用于语义线检测的数据集。语义线是有意义的线结构，勾勒出自然图像的概念结构。 NKL 数据集包含 5,000 张不同场景的图像。这些图像中的每一个都由多个熟练的人工注释者进行注释。数据集分为训练和验证子集。训练集中有 4,000 张图像，验证集中有 1,000 张图像。

NKL, the abbreviation for NanKai Lines, is a dataset designed for semantic line detection. Semantic lines are meaningful line structures that delineate the conceptual framework of natural images. The NKL dataset comprises 5,000 images covering diverse scenes. Each of these images has been annotated by multiple skilled human annotators. The dataset is split into training and validation subsets, with the training subset containing 4,000 images and the validation subset including 1,000 images.

搜集汇总

背景与挑战

背景概述

NKL数据集是一个用于语义线检测的数据集，包含5,000张不同场景的图像，分为4,000张训练图像和1,000张验证图像。该数据集由南开大学于2020年发布，适用于计算机视觉领域的语义线检测研究。