An extended set of Haar-like features for rapid object detection

By Rainer Lienhart, Jochen Maydt. 2002

3,464 citations recorded by the source corpus.

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

You Only Look Once: Unified, Real-Time Object Detection (2015, methodology): “Then, classifiers [35, 21, 13, 10] or localizers [1, 31] are used to identify objects in the feature space.”
Joint Haar-like features for face detection (2005, methodology): “Lienhart et al. [12] introduced an efficient scheme for calculating 45◦ rotated features.”
Efficient HOG human detection (2011, neutral): “To improve the representative and discriminating capacities, the original Haar-like rectangles have been extended to rotated features [17], diagonal features [17], and center-surrounded features [14].”
Towards semantic maps for mobile robots (2008, methodology): “The base resolution of the object detector is, e.g., 30 × 30 pixels, thus, the set of possible features in this area is very large (642,592 features, see [24] for calculation details).”
Recent Advances in Deep Learning for Object Detection (2019, methodology): “This fea- 39 ture vector was commonly encoded by low-level visual descriptors 40 such as SIFT (Scale Invariant Feature Transform) [21] , Haar [22] , 41 HOG (Histogram of Gradients) [19] or SURF (Speeded Up Robust 42 Features) [23] , which showed a certain robustness to scale, illumi- 43 nation and r”
Multiple Object Tracking Performance Metrics and Evaluation in a Smart Room Environment (2006, background): “When several types of trackers (upper body and full body) are available for the same person, a combined output is produced.”
Traffic Sign Recognition via Multi-Modal Tree-Structure Embedded Multi-Task Learning (2017, background): “…is a rather challenging task for intelligent transportation systems since signs in different subsets, e.g., speed limit signs, prohibition signs, and mandatory signs, are very different from each other in color or shape, whereas they share some similarities to the ones in the same subset.”
Facial feature detection using Haar classifiers (2006, methodology): “Since a frame rate of 5 frames per second was achieved in facial detection only by [5] using a much faster processor, regionalization provides a tremendous increase in efficiency in facial feature detection.”
A UAV Search and Rescue Scenario with Human Body Detection and Geolocalization (2007, background): “The work was a basis for several improvements, one of which was presented in [9].”
Feature representation for statistical-learning-based object detection: A review (2015, background): “such as rotated Haar-like features [39], sparse granule features [40].”
Robot task planning and explanation in open and uncertain worlds (2017, methodology): “For each view acquired, the OpenCV face detector (using cascades of simple detectors for simple Haar-like features [28]) is run to yield a number of potential faces.”
The Rat Grimace Scale: A partially automated method for quantifying pain in the laboratory rat via facial expressions (2011, methodology): “RFF detects rodent eyes and ears using boosted cascades of Haar classifiers [14], which use differences between pixel intensities in small rectangular regions (Haar-like features) to capture textural and orientation information, and combine the response from many such regions to make predictions on ”
Learning Modulated Loss for Rotated Object Detection (2019, background): “Object detection is an important and fundamental task in computer vision, and over the decades it has experienced a switch from traditional machine learning methods [1, 2, 3, 4] to deep learning methods.”
FloatBoost learning and statistical face detection (2004, methodology): “A coarse-to-fine view-partition strategy is used and this leads to a detector-pyramid architecture consisting of several levels from the coarse level on the top to the fine level at the bottom.”
A survey on face detection in the wild: Past, present and future (2015, background): “5, the feature set of [17] was generalized in [63] by introducing 45 degree rotated rectangular features (a-d), and center-surround features (e-f).”