Edge and Curve Detection for Visual Scene Analysis

By Azriel Rosenfeld, M. Thurston. 1971

Subfields: Computer Vision and Pattern Recognition

How researchers see this paper

The researchers whose trusted networks rank it highest, strongest advocate first. Every researcher gets a different order.

In Computer Vision and Pattern Recognition

Citations in the corpus

Soft-NMS — Improving Object Detection with One Line of Code (2017, methodology): “It was first employed in edge detection techniques [25].”
Multiscale Vision Transformers (2021, methodology): “In a parallel development, the computer vision community developed multiscale processing, sometimes called “pyramid” strategies, with Rosenfeld and Thurston [85], Burt and Adelson [8], Koenderink [61], among the key papers.”
Neural Motifs: Scene Graph Parsing with Global Context (2017, methodology): “We use non-maximal suppression (NMS) [35] at 0.3 IOU on the class level to pass 64 proposals to object context branch of our model.”
Unbiased Scene Graph Generation From Biased Training (2020, neutral): “For SGDet, 80 RoIs were sampled for each image and Per-Class NMS [47, 69] with 0.”
Retrieve, Read, Rerank: Towards End-to-End Multi-Document Reading Comprehension (2019, methodology): “Inspired by the non-maximum suppression (NMS) algorithm (Rosenfeld and Thurston, 1971) that is used to prune redundant bounding boxes in object detection (Ren et al.”
Scale-space filtering: A new approach to multi-scale description (1984, background): “This line of thinking has led to considerable interest in multi-scale descriptions [12],[2],[9] , [81.”
Color image segmentation: advances and prospects (2001, result): “The parallel edge detection technique implies that the decision of whether a set of points are on an edge or not is dependent on the gray level of the set and some set of its neighbors, which includes high-emphasis spatial frequency "ltering, gradient operators, adaptive local operator [9,10], funct”
Adaptive NMS: Refining Pedestrian Detection in a Crowd (2019, methodology): “It is an essential component of many detection methods, such as edge detection [33], feature point detection [25] and object detection [32, 20, 21].”
Textural Features Corresponding to Visual Perception (1978, background): “Unfortunately, since the current texture edge detectors [7]J[11] work successfully only between regions with different average gray-levels, it seems that these textural features are not directly usable for the region segmentation problem.”
Bounding Box Regression With Uncertainty for Accurate Object Detection (2018, methodology): “It is widely used in edge detection [44], feature point detection [37] and objection detection [13, 12, 42, 45].”
A Computational Approach to Edge Detection (1986, unknown): “Interestingly, Rosenfeld and Thurston [25] proposed exactly the opposite criterion in the choice of operator for edge detection in texture.”
BEVDet: High-performance Multi-camera 3D Object Detection in Bird-Eye-View (2021, background): “The other categories with relatively large sizes also beneﬁt from Scale-NMS Resolution The resolution of the signal channel is vital for BEVDet.”
Scaling Vision Transformers to Gigapixel Images via Hierarchical Self-Supervised Learning (2022, background): “Motivated by multiscale, pyramid-based image processing [11,42,61], recent progress in ViT architecture development has focused on efficiency and integration of multiscale information (e.”
Scale-MAE: A Scale-Aware Masked Autoencoder for Multiscale Geospatial Representation Learning (2022, background): “These include spatial pyramids [6, 33, 35, 48] and dense sampling of windows [32, 59, 60].”