SURF: Speeded Up Robust Features
Open accessThe DOI matches its Crossref record; 43 of the 43 references it cites were found in a registry or an index; none of the 18 sources its claims cite is checked for support yet.
Overview
CoChat Summary
TL;DR
SURF is a scale- and rotation-invariant interest point detector and descriptor that uses integral images, Hessian-based detection, and Haar-wavelet responses to be faster than prior methods while matching or exceeding their repeatability and distinctiveness. It reports strong results on standard benchmarks and a museum object recognition task.
Key findings
- The Fast-Hessian detector is more than 3× faster than DoG and 5× faster than Hessian-Laplace, with comparable or better repeatability on Graffiti, Leuven, Boat, and Wall; on the first Graffiti image (800×640), computation times were 120 ms for Fast-Hessian, 400 ms for DoG, 650 ms for Hessian-Laplace, and 1800 ms for Harris-Laplace.
- SURF’s 64-dimensional descriptor outperformed GLOH, SIFT, and PCA-SIFT in most comparisons, sometimes by more than 10% improvement in recall for the same precision; SURF-128 was slightly better but slower to match.
- Joint detector-descriptor computation times on the first Graffiti image were 255 ms for U-SURF, 354 ms for SURF, 391 ms for SURF-128, and 1036 ms for SIFT.
- In museum object recognition, recognition rates were 85.7% for SURF-128, 83.8% for U-SURF, 82.6% for SURF, 78.3% for GLOH, 78.1% for SIFT, and 72.3% for PCA-SIFT.
Methods
- Design: a scale- and rotation-invariant detector/descriptor. The detector uses the determinant of the Hessian with box-filter approximations evaluated via integral images; the descriptor uses Haar-wavelet responses in a 4×4 subregion grid, producing a 64-dimensional vector, with an upright U-SURF variant and an extended SURF-128 variant.
- Data: standard evaluation sequences from Mikolajczyk (including Graffiti, Wall, Boat, Leuven, Bikes, Trees, and Ubc; Bark excluded from descriptor evaluations) and a real-life museum object recognition database of 216 images of 22 objects, with 116 test images at 320×240.
- Analysis: detector repeatability scores, descriptor recall-(1-precision) graphs under similarity-threshold and nearest-neighbour-ratio matching, recognition rates, and computation timings on a standard Linux PC (Pentium IV, 3 GHz).
Limitations
- The paper has no explicit limitations section. It states that space limitations prevent showing results on all sequences, and it assumes skew, anisotropic scaling, and perspective effects are second-order effects covered only by overall descriptor robustness.
- Evident from the text: the method is scale- and rotation-invariant but not affine-invariant; Fig. 4 notes the interest points are not affine invariant, and the descriptor does not use colour.
- Evident from the text: the object recognition experiment uses small 320×240 test images and matching without additional geometric constraints, which the authors say would reduce false positives but were omitted for comparison purposes.
Model-written summary. Check it against the paper before relying on it.
Integrity signals
Registry
Cites
Statistics
Figures








Verification
Claims the paper makes
0 Valid · 2 Not valid · 12 Not checked · 30 Own results and interpretations
A result the paper reports from its own data — it cites no other source for it, so there is none to check. The quote shows it is in the paper.
In the paper
“The number of interest points found is on average very similar for all detectors.”
plain- Asserted: passes — The authors put it forward as true.
- Checkable: passes — It can be checked against evidence.
- Load-bearing: unclear — It is unclear whether the paper’s conclusions depend on it.
- Based: passes — The paper gives a reason why it holds.
A result the paper reports from its own data — it cites no other source for it, so there is none to check. The quote shows it is in the paper.
In the paper
“The 4 × 4 sub-region division solution provided the best results.”
plainstates a cause- Asserted: passes — The authors put it forward as true.
- Checkable: passes — It can be checked against evidence.
- Load-bearing: passes — The paper’s conclusions depend on it.
- Based: passes — The paper gives a reason why it holds.
A result the paper reports from its own data — it cites no other source for it, so there is none to check. The quote shows it is in the paper.
In the paper
“SURF outperformed the other descriptors for almost all the comparisons.”
plainstates a cause- Asserted: passes — The authors put it forward as true.
- Checkable: passes — It can be checked against evidence.
- Load-bearing: passes — The paper’s conclusions depend on it.
- Based: passes — The paper gives a reason why it holds.
A result the paper reports from its own data — it cites no other source for it, so there is none to check. The quote shows it is in the paper.
In the paper
“The SURF descriptor outperforms the other descriptors in a systematic and significant way, with sometimes more than 10% improvement in recall for the same level of precision.”
plainstates a cause- Asserted: passes — The authors put it forward as true.
- Checkable: passes — It can be checked against evidence.
- Load-bearing: passes — The paper’s conclusions depend on it.
- Based: passes — The paper gives a reason why it holds.
A result the paper reports from its own data — it cites no other source for it, so there is none to check. The quote shows it is in the paper.
In the paper
“As shown in the results section, the performance is comparable to the one using the discretised and cropped Gaussians.”
plain- Asserted: passes — The authors put it forward as true.
- Checkable: passes — It can be checked against evidence.
- Load-bearing: passes — The paper’s conclusions depend on it.
- Based: passes — The paper gives a reason why it holds.
8 more claims
Create a free account to see all 44 claims and every source verdict.
- It may state two separate claims as one. 24
- Probably not a substantial claim. 13
- The quote only partly states this claim. 1
Claims other papers make about this one
Each is checked against this paper’s own text: does it say what the citing paper says it does?
In the citing paper
“This transition marked a new era, where we moved beyond the classic approaches and task-specific feature engineering of earlier days [Lowe 2004; Bay et al. 2006; Rosten and Drummond 2006] towards models that could be trained once over large amounts of data, and then adapted for a broad variety of tasks, such as image recognition, object detection, and image segmentation [Krizhevsky et al. 2012; Sz”
From On the Opportunities and Risks of Foundation Models (2021) · cites this paper as Bay et al. 2006
Citations
Cites
Distinctive Image Features from Scale-Invariant Keypoints
Lowe, D. G. (2004) · International Journal of Computer Vision · doi:10.1023/b:visi.0000029664.99615.94 · Cited by 55,589
Rapid object detection using a boosted cascade of simple features
Viola, P., & Jones, M. (2005) · Proceedings of the 2001 IEEE Computer Society Conference on Computer Vision and Pattern Recognition. CVPR 2001 · doi:10.1109/cvpr.2001.990517 · Cited by 18,300
Object recognition from local scale-invariant features
Lowe, D. G. (1999) · Proceedings of the Seventh IEEE International Conference on Computer Vision · doi:10.1109/iccv.1999.790410 · Cited by 16,472
A Combined Corner and Edge Detector
Harris, C., & Stephens, M. (1988) · Procedings of the Alvey Vision Conference 1988 · doi:10.5244/c.2.23 · Cited by 12,901
A performance evaluation of local descriptors
Mikolajczyk, K., & Schmid, C. (2005) · IEEE Transactions on Pattern Analysis and Machine Intelligence · doi:10.1109/tpami.2005.188 · Cited by 6,723 · cited 14×
“This PCA-SIFT yields a 36-dimensional descriptor which is fast for matching, but proved to be less distinctive than SIFT in a second comparative study by Mikolajczyk et al. [8] and slower feature computation reduces the effect of fast matching.”
BackgroundMethodResultInfluentialMarr, D., & Hildreth, E. (1980) · Proceedings of the Royal Society of London. Series B. Biological Sciences · doi:10.1098/rspb.1980.0020 · Cited by 6,510
PCA-SIFT: a more distinctive representation for local image descriptors
Ke, Y., & Sukthankar, R. (2004) · Proceedings of the 2004 IEEE Computer Society Conference on Computer Vision and Pattern Recognition, 2004. CVPR 2004. · doi:10.1109/cvpr.2004.183 · Cited by 3,932
Robust wide-baseline stereo from maximally stable extremal regions
Matas, J., Chum, O., Urban, M., & Pajdla, T. (2004) · Image and Vision Computing · doi:10.1016/j.imavis.2004.02.006 · Cited by 3,686
Scalable Recognition with a Vocabulary Tree
Nister, D., & Stewenius, H. (2006) · 2006 IEEE Computer Society Conference on Computer Vision and Pattern Recognition - Volume 2 (CVPR'06) · doi:10.1109/cvpr.2006.264 · Cited by 3,579
Scale & Affine Invariant Interest Point Detectors
Mikolajczyk, K., & Schmid, C. (2004) · International Journal of Computer Vision · doi:10.1023/b:visi.0000027790.02288.f2 · Cited by 3,436
Cited by
Items 6,501–6,600 of 14,752
Search of objects of interest in videos
Mansencal, B., Benois‐Pineau, J., Vieux, R., et al. (2012) · International Workshop on Content-Based Multimedia Indexing (CBMI) · doi:10.1109/cbmi.2012.6269809 · Cited by 7
Real-time vehicle matching for multi-camera tunnel surveillance
Jelača, V., Castaneda, J. O. N., Frías-Velázquez, A., et al. (2011) · Proceedings of SPIE, the International Society for Optical Engineering/Proceedings of SPIE · doi:10.1117/12.876671 · Cited by 7
Image georegistration methods: A framework for application guidelines
Doucette, P. J., Antonisse, J., Braun, A., et al. (2013) · IEEE Applied Imagery Pattern Recognition Workshop · doi:10.1109/aipr.2013.6749317 · Cited by 7
Satellite Image Matching and Registration: A Comparative Study Using Invariant Local Features
Tahoun, M., Shabayek, A. E. R., Nassar, H., Giovenco, M., Reulke, R., Emary, E., & Hassanien, A. E. (2016) · Studies in computational intelligence · doi:10.1007/978-3-319-28854-3_6 · Cited by 7
ACTIVE STEREO-MATCHING FOR ONE-SHOT DENSE RECONSTRUCTION
Fernandez, S., Forest, J., & Salví, J. (2012) · International Conference on Pattern Recognition Applications and Methods · doi:10.5220/0003737005410545 · Cited by 7
Humanoid localization and navigation using a visual memory
Delfin, J., Becerra, H. M., & Arechavaleta, G. (2016) · IEEE-RAS International Conference on Humanoid Robots (Humanoids) · doi:10.1109/humanoids.2016.7803354 · Cited by 7
Conditional marked point process-based crowd counting in sparsely and moderately crowded scenes
Yoon, Y., Gwak, J., Song, J., et al. (2016) · International Conference on Control, Automation and Information Sciences (ICCAIS) · doi:10.1109/iccais.2016.7822463 · Cited by 7
Indoor calibration for stereoscopic camera STC: a new method
Dami, M., Veltroni, I. F., Cremonese, G., et al. (2017) · International Conference on Space Optics (ICSO) · doi:10.1117/12.2304188 · Cited by 7
A Visual Localization Technique for Unmanned Ground and Aerial Robots
Shim, J. H., & Cho, Y. I. (2017) · IEEE International Conference on Robotic Computing (IRC) · doi:10.1109/irc.2017.71 · Cited by 7
Ma, Y., & Shi, L. S. (2017) · 2017 International Applied Computational Electromagnetics Society Symposium (ACES) · Cited by 7
Bag of feature approach for vehicle classification in heterogeneous traffic
Das, J., Shah, M., & Mary, L. (2017) · IEEE International Conference on Signal Processing, Informatics, Communication and Energy Systems (SPICES) · doi:10.1109/spices.2017.8091346 · Cited by 7
“Bay, in his study [4] found that Hessian based detectors are more stable and repeatable than their Harris based counterparts.”
BackgroundMethodInfluentialPharmaPack: Mobile fine-grained recognition of pharma packages
Taran, O., Rezaeifar, S., Dabrowski, O., Schlechten, J., Holotyak, T., & Voloshynovskiy, S. (2017) · European Signal Processing Conference (EUSIPCO) · doi:10.23919/eusipco.2017.8081543 · Cited by 8
“The high dimensional feature extraction is based on either the usage of the last layers of deep nets trained in unsupervised or supervised way, a.k.a. neural codes [5], or aggregation of low- or- mid-dimensional local descriptors such as SIFT [6], SURF [7], aKaZe [8], etc. using feature aggregation such as Fisher vect…”
BackgroundClassification of Indian Monuments into Architectural Styles
Sharma, S., Aggarwal, P., Bhattacharyya, A. N., & Indu, S. (2018) · Communications in computer and information science · doi:10.1007/978-981-13-0020-2_47 · Cited by 8
An Improved Loop Closure Detection for RatSLAM
Gu, T., & Yan, R. (2019) · International Conference on Control, Automation and Robotics (ICCAR) · doi:10.1109/iccar.2019.8813378 · Cited by 7
“Over the recent years, binary features like oriented features from accelerated segment test and rotated binary robust independent elementary(ORB) [4] or LATCH [5], are gradually taking over the classical real-valued features such as scale-invariant feature transforms(SIFT) [6] or speeded-up robust features(SURF) [7] o…”
BackgroundImage Registration Method Based on Optimized SURF Algorithm
Sheng, Z., Peihua, L., Liu, Y., et al. (2019) · American Journal of Optics and Photonics · doi:10.11648/j.ajop.20190704.11 · Cited by 7
Performance Evaluation of Deep Neural Networks in Detecting Loop Closure of Visual SLAM
Zuo, L., Zhang, C. H., Liu, F. L., & Wu, Y. F. (2019) · International Conference on Intelligent Human-Machine Systems and Cybernetics · doi:10.1109/ihmsc.2019.10136 · Cited by 9
“First, the algorithm (SIFT [7], SURF [8], or ORB [9]) is used to extract the visual lexical vectors from the images of different classes.”
MethodRobust Feature Descriptors for Efficient Vision-Based Tracking
Carrera, G., Savage, J., & Mayol-Cuevas, W. W. (2007) · Springer eBooks · doi:10.1007/978-3-540-76725-1_27 · Cited by 7
“The descriptor is turned into a unit vector to achieve invariance to contrast [6].”
BackgroundMethodInfluentialEfficient region of interest detection using blind image division
Olaode, A. A., Naghdy, G. A., & Todd, C. (2015) · Signal Processing Symposium (SPSympo) · doi:10.1109/sps.2015.7168286 · Cited by 7
Guo, H., Xiao, H., Wang, S., et al. (2015) · IEEE International Conference on Mechatronics and Automation (ICMA) · doi:10.1109/icma.2015.7237740 · Cited by 7
“A very popular approach to detecting and recognizing objects is to establish correspondences between visual features in the input image and those in a database image [5-8].”
BackgroundGlobally optimized multiview video color correction using dense spatio-temporal matching
Ceulemans, B., Lu, S.-P., Schelkens, P., et al. (2015) · 3DTV-Conference: The True Vision - Capture, Transmission and Display of 3D Video · doi:10.1109/3dtv.2015.7169356 · Cited by 7
Robot localization using omnidirectional vision in large and dynamic outdoor environments
Ascani, A., Frontoni, E., Mancini, A., et al. (2008) · doi:10.1109/mesa.2008.4735695 · Cited by 7
Adaptive and optimal difference operators in image processing
Veelaert, P., & Teelen, K. (2008) · Pattern Recognition · doi:10.1016/j.patcog.2008.11.017 · Cited by 7
Target motion compensation with optical flow clustering during visual tracking
Huang, C., & Hung, M.-H. (2014) · IEEE International Conference on Networking, Sensing and Control (ICNSC) · doi:10.1109/icnsc.2014.6819607 · Cited by 7
Design and optimization of image processing algorithms on mobile GPU
Singhal, N., Yoo, J. W., Choi, H. Y., et al. (2011) · ACM SIGGRAPH Posters · doi:10.1145/2037715.2037741 · Cited by 7
Building topographic subspace model with transfer learning for sparse representation
Liu, Y., Cheng, J., Xu, C., & Lu, H. (2010) · Neurocomputing · doi:10.1016/j.neucom.2009.11.041 · Cited by 7
Extraction of 2D Barcode Using Keypoint Selection and Line Detection
Leong, L. K., & Wang, Y. (2009) · Lecture notes in computer science · doi:10.1007/978-3-642-10467-1_73 · Cited by 7
Recent Advances towards a Robust, Automated Hurricane Damage Assessment from High-Resolution Images
Thomas, J., Kareem, A., & Bowyer, K. W. (2012) · doi:10.1061/9780784412626.069 · Cited by 7
Statistical Classification Based Fast Drivable Region Detection for Indoor Mobile Robot
Qu, S., & Cai, M. (2014) · International Journal of Humanoid Robotics · doi:10.1142/s0219843614500108 · Cited by 7
Affine image region detection and description
Vázquez-Martín, R., Marfil, R., & Bandera, A. J. (2010) · Journal of Physical Agents (JoPha) · doi:10.14198/jopha.2010.4.1.07 · Cited by 8
EXTERIOR ORIENTATION ESTIMATION OF OBLIQUE AERIAL IMAGERY USING VANISHING POINTS
Verykokou, S., & Ioannidis, C. (2016) · The international archives of the photogrammetry, remote sensing and spatial information sciences/International archives of the photogrammetry, remote sensing and spatial information sciences · doi:10.5194/isprs-archives-xli-b3-123-2016 · Cited by 7
“Several tests performed for this study using multi-view oblique aerial images showed that the matching of SIFT features leads to better results than the computationally less expensive matching of SURF features (Bay et al., 2006).”
MethodAugmented Reality Framework for the Socialization between Elderly People
Almeida, L., Menezes, P. J., & Dias, J. M. M. (2013) · IGI Global eBooks · doi:10.4018/978-1-4666-3986-7.ch023 · Cited by 7
An automatic filtering algorithm for SURF-based registration of remote sensing images
Anzid, H., Goïc, G. L., Bekkari, A., et al. (2017) · International Conference on Advanced Technologies for Signal and Image Processing · doi:10.1109/atsip.2017.8075560 · Cited by 7
“The registration method in this paper is based on SURF method which allows a fast extraction of robust keypoints and ensures detection of enough shared keypoint in different locations for matching step with respect to repeatability and an acceptable computational time [25].”
MethodVision-based 3-D object pick-and-place tasks of industrial manipulator
Luo, G.-Y., Cheng, M.-Y., & Chiang, C.-L. (2017) · International Automatic Control Conference · doi:10.1109/cacs.2017.8284250 · Cited by 7
Ohlsson, N., & Ståhl, M. (2013) · Chalmers Publication Library (Chalmers University of Technology) · Cited by 7
Automatic video genre categorization and event detection techniques on large-scale sports data
Zhang, N., Duan, L., Huang, Q., et al. (2010) · Proceedings of CASCON · doi:10.1145/1923947.1923977 · Cited by 7
Algorithms for Visual Maritime Surveillance with Rapidly Moving Camera
Fefilatyev, S. (2012) · Digital Commons - University of South Florida (University of South Florida) · Cited by 11
PRELIMINARY STEREO RECONSTRUCTION FROM APOLLO 15 METRIC CAMERA IMAGERY. M.
Broxton, J., Moratto, Z. M., Nefian, A., et al. (2009) · Lunar and Planetary Science Conference · Cited by 7
Pooling the Convolutional Layers in Deep ConvNets for Action RecognitionPreprint
Zhao, S., Liu, Y., Han, Y., & Hong, R. (2015) · arXiv (Cornell University) · doi:10.48550/arxiv.1511.02126 · Cited by 6
What is the right way to represent document images?Preprint
Csurka, G., Larlus, D., Gordo, A., & Almazán, J. (2016) · arXiv (Cornell University) · doi:10.48550/arxiv.1603.01076 · Cited by 28
New sparse representation methods; application to image compression and indexing
Salvatierra, J. Z. (2010) · Cited by 7
Smart Library: Identifying Books in a Library using Richly Supervised Deep Scene Text ReadingPreprint
Yang, X., He, D., Huang, W., Zhou, Z., Ororbia, A. G., Kifer, D., & Giles, C. L. (2016) · arXiv (Cornell University) · doi:10.48550/arxiv.1611.07385 · Cited by 9
A CAD system for early diagnosis of autism using different imaging modalities.
Ismail, M. (2016) · ThinkIR: The University of Louisville's Institutional Repository (University of Louisville) · doi:10.18297/etd/2599 · Cited by 7
A Performance Evaluation of Local Features for Image Based 3D ReconstructionPreprint
Fan, B., Kong, Q., Wang, X., et al. (2017) · arXiv (Cornell University) · doi:10.48550/arxiv.1712.05271 · Cited by 7
Matching Features without Descriptors: Implicitly Matched Interest PointsPreprint
Cieslewski, T., Bloesch, M., & Scaramuzza, D. (2018) · arXiv (Cornell University) · doi:10.48550/arxiv.1811.10681 · Cited by 13
“Subsequently, faster approximations for distinctiveness have been found, whether using first-order approximations [11,12], or convolutional filters such as the Laplacian of Gaussian or Difference of Gaussian [4,5].”
MethodSwapped Face Detection using Deep Learning and Subjective AssessmentPreprint
Ding, X., Raziei, Z., Larson, E. C., et al. (2019) · arXiv (Cornell University) · doi:10.48550/arxiv.1909.04217 · Cited by 7
“They used sped up robust features (SURF) [4] and Bag of Words (BoW) to create image features instead of using raw pixels.”
MethodLevy, J., Jackson, C. R., Haudenschild, C. C., Christensen, B. C., & Vaickus, L. J. (2020) · bioRxiv (Cold Spring Harbor Laboratory) · doi:10.1101/2020.03.22.002402 · Cited by 6
“Similarity can be defined as intensity-based or feature-based, where the former calculates a correlation-based pixel-wise or window-wise intensity measure between the source and target images, while the latter technique identifies objects of interest using algorithms such as SIFT (Scale Invariant Feature Transform) 15…”
MethodLevy, J., Davis, M. A., Chacko, R. S., Davis, M., Fu, L., Goel, T., Pamal, A., Nafi, I., Angirekula, A., Christensen, B. C., et al. (2022) · medRxiv · doi:10.1101/2022.05.06.22274781 · Cited by 7
A Performance comparison of feature detectors for planetary rover mapping and localization
Wan, W., Peng, M., Xing, Y., Wang, Y., Liu, Z., Di, K., Teng, B., Mao, X., Zhao, Q., Xin, X., et al. (2017) · The International Archives of the Photogrammetry, Remote Sensing and Spatial Information Sciences · doi:10.5194/isprs-archives-xlii-3-w1-149-2017 · Cited by 6
“SURF (Speeded Up Robust Features)( Bay et al., 2006) is a high-performance scale and rotation-invariant interest point detector inspired by SIFT.”
MethodA novel technique for raindrop detection on a car windshield using geometric-photometric model
Sugimoto, M., Kakiuchi, N., Ozaki, N., & Sugawara, R. (2012) · 2012 15th International IEEE Conference on Intelligent Transportation Systems · doi:10.1109/itsc.2012.6338613 · Cited by 6
Real-Time Stereo-Based Ocean Surface Mapping for Robotic Floating Platforms: Concept and Methodology
Greenberg, O., & Moshe, B. B. (2023) · Sensors · doi:10.3390/s23083857 · Cited by 6
“Without getting too deep into details, Sparse stereo matching finds many “feature points” in one image (using classic algorithms such as Harris [34], SIFT [35], SURF [36], etc., or more-innovative methodologies such as R2D2 [37] or DELF [38]), then uses the Epipolar Geometry to find the “best” corresponding points on…”
MethodComputer Vision Algorithms on a Raspberry Pi 4 for Automated Depalletizing
Greco, D., Fasihiany, M., Ranjbar, A. V., Masulli, F., Rovetta, S., & Cabri, A. (2024) · Algorithms · doi:10.3390/a17080363 · Cited by 12
Image stitching method for CMOS grayscale cameras in industrial applications
Liu, Q., Huo, J., Tang, X., et al. (2024) · Optics & Laser Technology · doi:10.1016/j.optlastec.2024.111874 · Cited by 6
Human identification through panoramic dental radiographs: a novel matching approach
Bozkurt, M. H., Karagöl, S., & Ömezli, M. M. (2025) · Knowledge and Information Systems · doi:10.1007/s10115-025-02353-1 · Cited by 6
Self-Supervised Foundation Model for Template Matching
Hristov, A., Dimov, D., & Nisheva-Pavlova, M. M. (2025) · Big Data and Cognitive Computing · doi:10.3390/bdcc9020038 · Cited by 6
Tian, Y., Wan, X., Zhang, S., Zuo, J., Shao, Y., Liu, B., & Yang, M. (2025) · IEEE Transactions on Geoscience and Remote Sensing · doi:10.1109/tgrs.2025.3547292 · Cited by 11
“Some of these methods are still commonly used in planetary rovers, such as scale-invariant feature transform (SIFT) [18] and speeded up robust features (SURFs) [19].”
Mu, K., Wang, W., Liu, H., Liang, L., & Zhang, S. (2025) · Remote Sensing · doi:10.3390/rs17061071 · Cited by 6
Kosugi, S. (2025) · IEEE Transactions on Circuits and Systems for Video Technology · doi:10.1109/tcsvt.2025.3573401 · Cited by 7
“…correspondences: Ironi et al. [27] employed discrete cosine transform coef-ficients, Charpiat et al. [28] and Gupta et al. [29] utilized the SURF descriptor [30], Chia et al. [31] and Liu et al. [32] relied on SIFT features [33], and Fang et al. [34] used both HOG [35] and DAISY [36] descriptors.”
bin Hussain, M. D. I., Katiyar, V., Nagai, M., et al. (2025) · IEEE Journal of Selected Topics in Applied Earth Observations and Remote Sensing · doi:10.1109/jstars.2025.3582238 · Cited by 6
“Popularlocalinvariantfeaturesusedinimageregistrationin-cludethescale-invariantfeaturetransform[16],speededup robustfeatures[17],orientedFASTandrotatedBRIEF[18], binaryrobustinvariantscalablekeypoints[19],andKAZE[20].”
Sandhu, S. K., & Bhatia, R. (2025) · SN Computer Science · doi:10.1007/s42979-025-04173-0 · Cited by 9
“SURF [2] is a feature extractor algorithm which operates like the SIFT algorithm but is faster than SIFT.”
Reconstructing People, Places, and Cameras
Müller, L., Choi, H., Zhang, A., Yi, B., Malik, J., & Kanazawa, A. (2025) · IEEE Computer Society Conference on Computer Vision and Pattern Recognition/Proceedings - IEEE Computer Society Conference on Computer Vision and Pattern Recognition/Proceedings · doi:10.1109/cvpr52734.2025.02044 · Cited by 31
“…and a global alignment loss that merges per-view pointmaps into the same world space. ods [1, 9, 10], such as COLMAP [41, 42] employs key-point detection, matching based on locally invariant descriptors [4, 30, 40], and incremental bundle adjustment to estimate camera poses and sparse 3D points.”
A scalable benchmark to evaluate the robustness of image stitching under simulated distortions
Liu, Y., Wang, Y., Wang, D., et al. (2025) · Scientific Reports · doi:10.1038/s41598-025-17730-7 · Cited by 6
“Feature-based stitching methods : Feature-based approaches, such as SIFT 12 , SURF 23 , ORB 24 , and AKAZE 14 , detect and match keypoints between overlapping images.”
Zhang, S., Luo, B., Liu, J., Fu, Z., Su, X., & Zhu, S. (2025) · IEEE Transactions on Geoscience and Remote Sensing · doi:10.1109/tgrs.2025.3599445 · Cited by 14
“Several classic algorithms, such as SIFT [36], SURF [53], BRISK [54], and ORB [40], perform well under linear intensity variations but are sensitive to NID, limiting their applicability to multimodal remote sensing images.”
EDM: Efficient Deep Feature Matching
Li, X., Rao, T., & Pan, C. (2025) · IEEE/CVF International Conference on Computer Vision (ICCV) · doi:10.1109/iccv51701.2025.02431 · Cited by 22
“[4, 5, 35, 50].”
CPFL: Resilient Continuous UAV Localization via Cross-View Perception and Particle Filtering
Su, C., Yuan, J., Enhui, Z., Xu, W., Liu, Z., & Hu, J. (2026) · Drones · doi:10.3390/drones10060437 · Cited by 21
A framework based on vision sensors for the automatic management of exchange parking areas
Catani, L., Frontoni, E., & Zingaretti, P. (2010) · doi:10.1109/mesa.2010.5552047 · Cited by 6
“The solution adopted is to perform the tracking of the vehicle of interest and to keep track of the position occupied in the area along with some other information such as size, color and features, for example, identified by applying the SURF (Speeded Up Robust Features) algorithm [10].”
MethodVintr, T., Pastorek, L., Vintrová, V., et al. (2011) · World Congress on Nature and Biologically Inspired Computing (NaBIC) · doi:10.1109/nabic.2011.6089625 · Cited by 6
License plate detection based on Speeded Up Robust Features and Bag of Words model
Khaleel, F. M., Abdullah, S. N. H. S., & Ismail, M. K. B. (2013) · IEEE International Conference on Smart Instrumentation, Measurement and Applications (ICSIMA) · doi:10.1109/icsima.2013.6717937 · Cited by 6
“Speeded Up Robust Features (SURF) technique was proposed to use for object detection [4].”
MethodLearning Compact Visual Descriptors for Low Bit Rate Mobile Landmark Search
Duan, L., Chen, J., Ji, R., et al. (2013) · AI Magazine · doi:10.1609/aimag.v34i2.2469 · Cited by 6
Deep learning object recognition in multi-spectral UAV imagery
Knyaz, V. A., & Zheltov, S. Y. (2018) · Optics, Photonics, and Digital Technologies for Imaging Applications · doi:10.1117/12.2307661 · Cited by 7
Rotation-Invariant Descriptor for Disparate Images Using Line Segments
Sahoo, P., Sharma, T., Agrawal, P., et al. (2018) · Advances in intelligent systems and computing · doi:10.1007/978-981-13-1135-2_30 · Cited by 6
Capponi, L., Tocci, T., D’Imperio, M., et al. (2021) · ACTA IMEKO · doi:10.21014/acta_imeko.v10i4.1148 · Cited by 6
Summarizing large scale 3D mesh for urban navigation
Salah, I. B., Kramm, S., Demonceaux, C., et al. (2022) · Robotics and Autonomous Systems · doi:10.1016/j.robot.2022.104037 · Cited by 6
“We mention among these algorithms, the SUSAN detector [45], the FAST detector[46], the SIFT detector [47, 48], the SURF detector [49], the MSER detector [50], the CenSurE [51] detector, the AGAST [52] detector and the Harris3D [53] detector which is known for its simplicity and efficiency in…”
MethodChen, H., Sun, D., Liu, W., et al. (2022) · IEEE Transactions on Geoscience and Remote Sensing · doi:10.1109/tgrs.2022.3175758 · Cited by 6
“Image feature detector and descriptor algorithms include scale-invariant feature transform (SIFT) [26], speeded up robust features (SURFs) [27], and Harris.”
MethodKim, G. Y., Kim, J. Y., Lee, S. H., et al. (2022) · BioMed Research International · doi:10.1155/2022/1705338 · Cited by 6
Lee, S., & Yeh, C.-H. (2022) · Journal of Ambient Intelligence and Smart Environments · doi:10.3233/ais-210129 · Cited by 6
“Image feature extraction technologies include SIFT (Scale-Invariant Feature Transform) [25], SURF (Speeded Up Robust Feature) [4] and GHT-based (General Hough Transform-based) [36] method.”
Matalov, D. P., Limonova, E. E., Skoryukina, N. S., et al. (2022) · IEEE Access · doi:10.1109/access.2022.3233463 · Cited by 6
“The algorithm, that used a combination of SURF [16] keypoints and RFD [18] descriptors, was claimed to find document coordinates and its class on iPhone 6 smart-phone for 0.35 seconds.”
MethodInfluentialBubble Trajectory Tracking Based on ORB Algorithm
Wang, S., Lu, S., Liu, J., et al. (2023) · Image Analysis & Stereology · doi:10.5566/ias.2794 · Cited by 6
Exploring the Role of Explainability in AI-Assisted Embryo Selection
Urcelay, L., Hinjos, D., Martin-Torres, P. A., González, M., Méndez, M., Cívico, S., Álvarez-Napagao, S., & García-Gasulla, D. (2023) · Frontiers in artificial intelligence and applications · doi:10.3233/faia230678 · Cited by 10
Deepa, R. N. A., Sankaranarayanan, S., Padthe, A., & Ramannavar, M. M. (2023) · International Journal of Advanced Computer Science and Applications · doi:10.14569/ijacsa.2023.0141036 · Cited by 7
Adaptive Pose Estimation Algorithm Based on Point Pair Feature
Chen, Y., Li, Q., & Zhang, M. (2024) · IEEE Sensors Journal · doi:10.1109/jsen.2024.3355923 · Cited by 6
“D descriptors include scale-invariant feature transform (SIFT) [17], features from accelerated segment test (FAST) [18], speeded-up robust features (SURF) [16], and oriented FAST and rotated BRIEF (ORB) [19].”
BackgroundImage registration for in situ X-ray nano-imaging of a composite battery cathode with deformation
Su, B., Qian, G., Gao, R., et al. (2024) · Journal of Synchrotron Radiation · doi:10.1107/s1600577524000146 · Cited by 6
Conv-Attention: A Low Computation Attention Calculation Method for Swin Transformer
Lou, Z., Luo, S., Huang, X., & Wei, D. (2024) · Neural Processing Letters · doi:10.1007/s11063-024-11483-6 · Cited by 7
“In early work, bag-of-words (BoW [26]) based models are widely used, which rely on hand-craft features (SIFT [22], SURF [3], HoG [8], Gist [23]).”
LIP-Loc: LiDAR Image Pretraining for Cross-Modal Localization
Puligilla, S. S., Omama, M., Zaidi, H., Parihar, U. S., & Krishna, K. M. (2024) · IEEE/CVF Winter Conference on Applications of Computer Vision Workshops (WACVW) · doi:10.1109/wacvw60836.2024.00103 · Cited by 58
“Recently, LiDARCLIP [14] uses principles from CLIP to produce joint text, LiDAR and camera embeddings, although not for the task of global localization.”
MethodMachine Vision Based Local Hyacinth Bean Breed Recognition Using Convolutional Neural Network
Khan, M. A. A., Rahman, M. A., Hossain, M. L., & Habib, M. T. (2024) · International Conference on Advances in Computing, Communication, Electrical, and Smart Systems (iCACCESS) · doi:10.1109/icaccess61735.2024.10499455 · Cited by 6
“Conventional feature descriptors including speeded up robust features (SURF) [14], scale invariant feature transform (SIFT) [13], local binary pattern (LBP) [11–12], and histogram of oriented gradients (HOG) [15] are utilized to get better classifications.”
MethodAn Efficient Multiple Convolutional Neural Network Model (MCNN-14) for Fashion Image Classification
Saed, S., Teimourpour, B., Kalashi, K., & Soltanshahi, M. A. (2024) · International Conference on Web Research (ICWR) · doi:10.1109/icwr61162.2024.10533341 · Cited by 11
“Common feature extraction techniques used in these early efforts included Histogram of Oriented Gradients (HOG) [17], Speeded Up Robust Features (SURF) [18], Scale-Invariant Feature Transform (SIFT) [19], and Features from Accelerated Segment Test (FAST) [20].”
MethodResearch on Deep Learning-Based Vehicle and Pedestrian Object Detection Algorithms
Zhang, X., Huang, H., Yang, J., & Jiang, S. (2024) · The international archives of the photogrammetry, remote sensing and spatial information sciences/International archives of the photogrammetry, remote sensing and spatial information sciences · doi:10.5194/isprs-archives-xlviii-4-w10-2024-213-2024 · Cited by 7
Dor, M., Driver, T., Getzandanner, K. M., & Tsiotras, P. (2024) · The International Journal of Robotics Research · doi:10.1177/02783649241234367 · Cited by 22
Huangfu, W., Ni, C., Wang, P., & Zhang, Y. (2024) · Applied Intelligence · doi:10.1007/s10489-024-05600-0 · Cited by 6
“Numerous well-known conventional feature matching techniques, such as FAST (features from accelerated segment test) [10], SIFT (scale-invariant feature transform) [11, 12], ORB (oriented FAST and rotated BRIEF) [13], and SURF (speeded up robust features) [14], have produced satisfactory results.”
Improved Feature Point Extraction Method of VSLAM in Low-Light Dynamic Environment
Wang, Y., Zhang, Y., Li-he, H., Ge, G., Wang, W., & Tan, S. (2024) · Electronics · doi:10.3390/electronics13152936 · Cited by 6
Gong, S., Long, Y., Chen, K., Liu, J., Xiao, Y., Chen, L., Wang, Z., & Dou, Q. (2024) · IEEE Transactions on Medical Imaging · doi:10.1109/tmi.2024.3439701 · Cited by 11
“Some works propose extracting feature descriptors like SIFT [27], SURF [28], or ORB [12] and then recovering the dense deformation field through interpolation.”
Intelligent Sensing Framework: Near-Sensor Machine Learning for Efficient Data Transmission
Huang, W. M., Rezvani, A., Chen, H., Ni, Y., Yun, S., Jeong, S., Zhang, G., & Imani, M. (2024) · IEEE Sensors Journal · doi:10.1109/jsen.2024.3440988 · Cited by 16
“Depending on the specific scenario in which it is being applied, ATC approaches utilize a variety of traditional feature extraction algorithms, ranging from handcrafted methods (e.g., [24], [25], [26]) to information theory-based methods [27], [28].”
MethodAhmed, S., Girerd, T., Clare, A. T., et al. (2024) · Journal of Materials Processing Technology · doi:10.1016/j.jmatprotec.2024.118571 · Cited by 6
CICA: Content-Injected Contrastive Alignment for Zero-Shot Document Image Classification
Sinha, S., Khan, M. S. U., Sheikh, T. U., Stricker, D., & Afzal, M. Z. (2024) · Lecture notes in computer science · doi:10.1007/978-3-031-70546-5_8 · Cited by 11
Yang, Z., Pu, F., Chen, H., et al. (2024) · IEEE Journal of Selected Topics in Applied Earth Observations and Remote Sensing · doi:10.1109/jstars.2024.3493883 · Cited by 6
“Traditional handcrafted feature point extraction, such as SIFT [23], SURF [24], KAZE [25], and ORB [26], rely on the local structure and pattern of the image, such as gradient and intensity changes.”
Wu, P., Lu, B., Li, H., et al. (2024) · Computer-Aided Civil and Infrastructure Engineering · doi:10.1111/mice.13383 · Cited by 6
AMPLE: Automatic Progressive Learning for Orientation Unknown Ground-to-Aerial Geo-Localization
Li, C. C., Yan, C., Xiang, X., Lai, J., Zhou, H., & Tang, D. (2024) · IEEE Transactions on Geoscience and Remote Sensing · doi:10.1109/tgrs.2024.3517654 · Cited by 10
“In addressing ground-to-aerial geo-localization tasks, conventional approaches initially relied on manually designed features to tackle the challenge [4], [5], [6].”
Uncertainty Guided Deep Lucas-Kanade Homography for Multimodal Image Alignment
Zhou, Z., Luo, J., Zhu, Q. X., Wang, Y., Zhong, H., Feng, M., & Chen, L. (2024) · IEEE Transactions on Geoscience and Remote Sensing · doi:10.1109/tgrs.2024.3522387 · Cited by 6
Zeng, Q., Hui, B., Liu, Z., et al. (2025) · Sensors · doi:10.3390/s25051345 · Cited by 6
“Traditional feature-based matching methods typically involve first detecting features, followed by a feature description [2,10–12].”
Toward Specialized Learning-based Approaches for Visual Odometry: A Comprehensive Survey
Phan, T.-D., & Kim, G.-W. (2025) · Journal of Intelligent & Robotic Systems · doi:10.1007/s10846-025-02245-0 · Cited by 11
“Conventionally, hand-crafted feature extractors [26–28] have been employed in the selection of specific image features or descriptors deemed pertinent to a given task.”
Liao, Y., Su, J., Ma, D., & Niu, C. (2025) · Remote Sensing · doi:10.3390/rs17142448 · Cited by 7
Metadata
- Authors
- Bay, H., Tuytelaars, T., & Van Gool, L.
- Affiliations
- ETH Zurich (CH)
- KU Leuven (BE)
- ORCID
- Tuytelaars, Tinne
- Venue
- Lecture Notes in Computer Science
- Publisher
- Springer Berlin Heidelberg
- Published
- 2006
- Pages
- 404-417
- ISSN
- 0302-9743 (ISSN-L)
- 1611-3349
- OpenAlex
- W1677409904
- Semantic Scholar
- 9ad3179098eb9ac7f0c81bf89b6bdcde91752880
- Cited by
- 14,955
- References
- 39
- Citations by year
- 2026: 210 · 2025: 519 · 2024: 727 · 2023: 757 · 2022: 878 · 2021: 975 · 2020: 978 · 2019: 1036 · 2018: 968 · 2017: 976 · 2016: 930 · 2015: 1079 · 2014: 1006 · 2013: 982 · 2012: 884
- FWCI
- 196.80
- Type
- conference paper
- Language
- en
- Open access
- bronze
- doi.org
- Licence
- www.springer.com/tdm (tdm)
- Locations
- Lecture notes in computer science (publishedVersion)
- Lirias (KU Leuven) (submittedVersion, other-oa)
- Funders
- Schweizerischer Nationalfonds zur Förderung der Wissenschaftlichen Forschung
- Topics
- Physical Sciences › Computer Science › Computer Vision and Pattern Recognition › Advanced Image and Video Retrieval Techniques
- Physical Sciences › Computer Science › Computer Vision and Pattern Recognition › Image Retrieval and Classification Techniques
- Physical Sciences › Computer Science › Computer Vision and Pattern Recognition › Image and Object Detection Techniques
- Keywords
- SURF, keypoint detection, feature descriptors, scale invariance, rotation invariance, integral image, Hessian matrix, feature matching, object recognition, repeatability
About these checks
Registry, retraction, Cites and statistics results are computed, not judged, with one exception: a printed reference that no index lists is read once by a model before it counts as a citation error. Claims are extracted by a model and kept only when their quote appears verbatim in the text; each cited source is then checked, with a model’s help, for whether it says what the claim relies on. Those verdicts can be wrong — read the claim’s quote and the cited source before relying on one.
A language model drew the claims from the paper on 2026-10-10, under version 3 of the claim definition (model deepseek-v4.1-flash, prompt extract-v3).
A second model labelled each claim on 2026-10-10 (model jev-1.13-20260917, definition revision gold-guideline-v0.2).
Statistics and arithmetic are recomputed, not judged.
Confidence bands are shown only for fields calibrated on our annotated test set. None is calibrated yet, so no band is shown.
What counts as a claim ›How we measure confidence ›How we check this ›
Are you an author, or do you see a problem with this page? Request a review
Data sources
Bibliographic data: Semantic Scholar, OpenAlex, Crossref.
Contains information from the Semantic Scholar Academic Graph Datasets, made available under the ODC Attribution License (ODC-BY 1.0).
Semantic Scholar’s data is described in: Kinney, R., et al. (2023). The Semantic Scholar Open Data Platform. arXiv:2301.10140.