Fast Edge Detection Using Structured Forests

By Piotr Doll'ar, C. Lawrence Zitnick. 2014

1,006 citations recorded by the source corpus.

Edge detection is a critical component of many vision systems, including object detectors and image segmentation algorithms. Patches of edges exhibit well-known forms of local structure, such as straight lines or T-junctions. In this paper we take advantage of the structure present in local image patches to learn both an accurate and computationally efficient edge detector. We formulate the problem of predicting local edge masks in a structured learning framework applied to random decision forests. Our novel approach to learning decision trees robustly maps the structured labels to a discrete space on which standard information gain measures may be evaluated. The result is an approach that obtains realtime performance that is orders of magnitude faster than many competing state-of-the-art approaches, while also achieving state-of-the-art edge detection results on the BSDS500 Segmentation dataset and NYU Depth dataset. Finally, we show the potential of our approach as a general purpose edge detector by showing our learned edge models generalize well across datasets.

Subfields: Computer Vision and Pattern Recognition, Media Technology

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

In Media Technology

Citations in the corpus

InstanceCut: From Edges to Instances with MultiCut (2016, background): “An alternative approach suggests to treat the problem as a per-pixel classiﬁcation task [35, 17].”
What Makes for Effective Detection Proposals? (2015, methodology): “ments is typically based on a diverse set of cues including superpixel shape, appearance cues, and boundary estimates (typically obtained from [35], [36]).”
EdgeConnect: Generative Image Inpainting with Adversarial Edge Learning (2019, background): “Doll ´ ar and Zitnikc [11] use structured learning [35] on random decision forests to predict local edge masks.”
EDTER: Edge Detection with Transformer (2022, methodology): “[49], gPb+NG [19], SE [15], SE+NG+ [20], OEF [21], SemiContour [69], HED [65], RCF [36], AMH-Net [66], LPCB [12], BDCN [22], and PiDiNet [53].”
Holistically-Nested Edge Detection (2015, methodology): “ing, one may categorize works into a few groups such as I: early pioneering methods like the Sobel detector [20], zerocrossing [27, 37], and the widely adopted Canny detector [4]; methods driven by II: information theory on top of features arrived at through careful manual design, such as Statistica”
Learning Rich Features from RGB-D Images for Object Detection and Segmentation (2014, methodology): “We can also use our cues with [10], and observe an analogous improvement in performance (Our(SE+SH + all cues) [RGBD] dashed blue curve in Figure 3).”
UberNet: Training a Universal Convolutional Neural Network for Low-, Mid-, and High-Level Vision Using Diverse Datasets and Limited Memory (2016, neutral): “he VOC context dataset [71] and evaluate it on the test set of the Berkeley Segmentation Dataset (BSD) of [69] and. We compare our method to some of the best-established methods of boundary detection [1, 23], as well as more recent, deep learning-based ones [6,28,41,48,49,90,107]. Method ODS OIS AP ”
DeepEdge: A multi-scale bifurcated deep network for top-down contour detection (2014, methodology): “Some recent local methods include sketch tokens [16] and structured edges [9], Both of these methods are trained in a supervised fashion using a random forest classiﬁer.”
Reading Text in the Wild with Convolutional Neural Networks (2014, methodology): “Following [62], we compute the edge response map using the Structured Edge detector [15, 16] and perform Non-Maximal Suppression orthogonal to the edge responses, sparsifying the edge map.”
Unsupervised Learning of Edges (2015, methodology): “Illustration of edge detection results on five sample images (same as used in [11]).”
High-for-Low and Low-for-High: Efficient Boundary Detection from Deep Object Features and Its Applications to High-Level Vision (2015, methodology): “These examples illustrate the effective advantage that HFL boundaries provide over another state-of-the-art edge detection system, the SE system [6].”
Oriented edge forests for boundary detection (2014, methodology): “In order to provide a more detailed prediction of the contour shape, we utilize a local segmentation procedure similar to the sharpening method introduced by Dollár and Zitnick [10].”
Weakly Supervised Instance Segmentation Using Class Peak Response (2018, methodology): “Multi-scale Combinatorial Grouping (MCG) [27] uses low-level contour information, e.g ., Structured Edge [8] or Ultrametric Contour Map [20], to extract object proposals, which contain ﬁne-detailed object boundaries that is valuable to instance segmentation.”
Learning Deep Structured Multi-Scale Features using Attention-Gated CRFs for Contour Prediction (2017, methodology): “Method ODS OIS AP gPb-ucm [1] .632 .661 .562 OEF [15] .651 .667 – Silberman et al. [33] .658 .661 – SemiContour [46] .680 .700 .690 SE [10] .685 .699 .679 gPb+NG [13] .687 .716 .629 SE+NG+ [14] .710 .723 .738 HED (RGB) [38] .720 .734 .734 HED (HHA) [38] .682 .695 .702 HED (RGB + HHA) [38] .746 .761…”
Joint Bi-layer Optimization for Single-Image Rain Streak Removal (2017, methodology): “The first application is edge detection using a recent edge detector [6].”
Richer Convolutional Features for Edge Detection (2016, methodology): “These methods, such as Pb [8], gPb [9], and SE [10], usually employ”