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 701–800 of 14,752
An accurate finger vein based verification system
Gupta, P., Gupta, P., Gupta, P., et al. (2014) · Digital Signal Processing · doi:10.1016/j.dsp.2014.12.003 · Cited by 97
Causally Regularized Learning with Agnostic Data Selection Bias
Shen, Z., Cui, P., Kuang, K., Li, B., & Chen, P. (2018) · ACM International Conference on Multimedia (ACM MM) · doi:10.1145/3240508.3240577 · Cited by 119
Feng, R., Du, Q., Li, X., & Shen, H. (2019) · ISPRS Journal of Photogrammetry and Remote Sensing · doi:10.1016/j.isprsjprs.2019.03.002 · Cited by 98
Learning and Aggregating Deep Local Descriptors for Instance-Level Recognition
Tolias, G., Jenícek, T., & Chum, O. (2020) · Lecture notes in computer science · doi:10.1007/978-3-030-58452-8_27 · Cited by 128
“Numerous classical approaches that are based on hand-crafted local features [27,25] and descriptors [26,8] exist in the literature of instance-level search [42,32,44,51,30].”
BackgroundLocal feature matching using deep learning: A survey
Xu, S., Chen, S., Xu, R., Wang, C., Lu, P., & Guo, L. (2024) · Information Fusion · doi:10.1016/j.inffus.2024.102344 · Cited by 155
“Additionally, certain algorithms have been devised to cater to both feature detection and description [33, 34, 35, 36, 37, 38].”
MethodDeep Learning for Visual Localization and Mapping: A Survey
Chen, C., Wang, B., Lu, C. X., Trigoni, N., & Markham, A. (2024) · IEEE Transactions on Neural Networks and Learning Systems · doi:10.1109/tnnls.2023.3309809 · Cited by 97
“The common pipeline varies from using handcrafted detectors [145], [146] and descriptors [147], [148], replacing either the descriptor [118], [149], [150], [151], [152], [153], [154], [155], [156], [157] or detector [158], [159], [160] with a learned alternative, or learning both the detector and…”
Transform coding of image feature descriptors
Chandrasekhar, V., Gabriel Takacs, A., Chen, D. M., Tsai, S. S., Singh, J., & Girod, B. (2008) · Proceedings of SPIE, the International Society for Optical Engineering/Proceedings of SPIE · doi:10.1117/12.805982 · Cited by 96
Comparison of 3D interest point detectors and descriptors for point cloud fusion
Hänsch, R., Weber, T., & Hellwich, O. (2014) · ISPRS annals of the photogrammetry, remote sensing and spatial information sciences · doi:10.5194/isprsannals-ii-3-57-2014 · Cited by 107
Quaternionic Local Ranking Binary Pattern: A Local Descriptor of Color Images
Lan, R., Zhou, Y., & Tang, Y. Y. (2015) · IEEE Transactions on Image Processing · doi:10.1109/tip.2015.2507404 · Cited by 95
“…include scale-invariant feature transform (SIFT) [10], gradient location-orientation histogram (GLOH) [32], speeded up robust features (SURF) [11], local binary pattern (LBP) [33], histograms of oriented gradient (HOG) descriptor [34], GA-SIFT [35], and Weber local descriptor (WLD) [36],…”
BackgroundImproving monocular visual SLAM in dynamic environments: an optical-flow-based approach
Cheng, J., Sun, Y., & Meng, M. Q. ‐. (2019) · Advanced Robotics · doi:10.1080/01691864.2019.1610060 · Cited by 103
“D feature points based on SURF descriptor and they use the IMU to compensate rotation of feature points to have the same rigid body transformation matrix between the successive images from the RGB-D sensor.”
Method3D Tree Dimensionality Assessment Using Photogrammetry and Small Unmanned Aerial Vehicles
Gatziolis, D., Liénard, J., Vogs, A., & Strigul, N. S. (2015) · PLoS ONE · doi:10.1371/journal.pone.0137765 · Cited by 109
Mouats, T., Aouf, N., Sappa, A. D., Aguilera, C., & Toledo, R. (2014) · IEEE Transactions on Intelligent Transportation Systems · doi:10.1109/tits.2014.2354731 · Cited by 96
Controlling UAVs with sensor input spoofing attacks
Davidson, D., Wu, H., Jellinek, R., Ristenpart, T., & Singh, V. (2016) · Cited by 173
Transfer Neural Trees for Heterogeneous Domain Adaptation
Chen, W.-Y., Hsu, T.-M. H., Tsai, Y.-H. H., Wang, Y.-C. F., & Chen⋆, M.-S. (2016) · Lecture notes in computer science · doi:10.1007/978-3-319-46454-1_25 · Cited by 105
“In Table 2, we compare the results of using NN, dNDF and our Transfer-NDF as G in TNT (with DeCAF6 and SURF describing source and target-domain data, respectively).”
MethodInfluentialHartmann, W., Havlena, M., & Konrad Schindler, B. (2014) · 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/cvpr.2014.9 · Cited by 104
Mixed reality and remote sensing application of unmanned aerial vehicle in fire and smoke detection
Esfahlani, S. S. (2019) · Journal of Industrial Information Integration · doi:10.1016/j.jii.2019.04.006 · Cited by 110
Yahyanejad, S., & Rinner, B. (2014) · ISPRS Journal of Photogrammetry and Remote Sensing · doi:10.1016/j.isprsjprs.2014.07.015 · Cited by 95
Zitnick, C. L., & Ramnath, K. (2011) · IEEE/CVF International Conference on Computer Vision (ICCV) · doi:10.1109/iccv.2011.6126263 · Cited by 103
Improving Descriptors for Fast Tree Matching by Optimal Linear Projection
Mikolajczyk, K., & Matas, J. (2007) · 2007 IEEE 11th International Conference on Computer Vision · doi:10.1109/iccv.2007.4408871 · Cited by 93
Learning Kernels for Unsupervised Domain Adaptation with Applications to Visual Object Recognition
Gong, B., Grauman, K. L., & Sha, F. (2014) · International Journal of Computer Vision · doi:10.1007/s11263-014-0718-4 · Cited by 100
“Briefly, we use SURF features (Bay et al. 2006) and encode the images with 800bin histograms with the codebook trained from a subset of Amazon images.”
MethodVineyard yield estimation by automatic 3D bunch modelling in field conditions
Herrero-Huerta, M., González‐Aguilera, D., Rodríguez‐Gonzálvez, P., & Hernández‐López, D. (2014) · Computers and Electronics in Agriculture · doi:10.1016/j.compag.2014.10.003 · Cited by 101
Locality-constrained and spatially regularized coding for scene categorization
Shabou, A., & Borgne, H. L. (2012) · 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/cvpr.2012.6248107 · Cited by 97
Semi-Supervised Zero-Shot Classification with Label Representation Learning
Li, X., Guo, Y., & Schuurmans, D. (2015) · 2015 IEEE International Conference on Computer Vision (ICCV) · doi:10.1109/iccv.2015.479 · Cited by 91
Ma, Z., Choi, J., & Sohn, H. (2021) · Computer-Aided Civil and Infrastructure Engineering · doi:10.1111/mice.12767 · Cited by 94
“Feature points, 𝑀 , 𝑀 , ... , are detected within the first ROI and matched with the corresponding features, 𝑀 , 𝑀 ,.... within the i th ROI using speeded up robust features (SURF) algorithm (Bay et al., 2006).”
MethodInfluentialComputer Vision-Based Carbohydrate Estimation for Type 1 Patients With Diabetes Using Smartphones
Anthimopoulos, M., Dehais, J., Shevchik, S., Ransford, B. H., Duke, D., Diem, P., & Mougiakakou, S. (2015) · Journal of Diabetes Science and Technology · doi:10.1177/1932296815580159 · Cited by 93
Pose estimation with radial distortion and unknown focal length
Josephson, K., & Byröd, M. (2009) · 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/cvpr.2009.5206756 · Cited by 98
Cascade of descriptors to detect and track objects across any network of cameras
Alahi, A., Vandergheynst, P., Bierlaire, M., & Kunt, M. (2010) · Computer Vision and Image Understanding · doi:10.1016/j.cviu.2010.01.004 · Cited by 92
Multi-Level Dense Descriptor and Hierarchical Feature Matching for Copy–Move Forgery Detection
Bi, X., Pun, C., & Yuan, X. (2016) · Information Sciences · doi:10.1016/j.ins.2016.01.061 · Cited by 97
Learning Video Stabilization Using Optical Flow
Yu, J., & Ramamoorthi, R. (2020) · 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/cvpr42600.2020.00818 · Cited by 101
“We compute SURF features[1] and their matches between consecutive frames, then compute the affine transformations.”
MethodDistracted driver detection by combining in-vehicle and image data using deep learning
Omerustaoglu, F., Sakar, C. O., & Kar, G. (2020) · Applied Soft Computing · doi:10.1016/j.asoc.2020.106657 · Cited by 104
RGB-D and Thermal Sensor Fusion: A Systematic Literature Review
Brenner, M., Reyes, N. H., Sušnjak, T., & Barczak, A. L. C. (2023) · IEEE Access · doi:10.1109/access.2023.3301119 · Cited by 141
SURF applied in panorama image stitching
Juan, L., & Oubong, G. (2010) · International Conference on Image Processing Theory, Tools and Applications (IPTA) · doi:10.1109/ipta.2010.5586723 · Cited by 94
“On the other hand, SURF creates a “stack” without 2:1 down sampling for higher levels in the pyramid, resulting in images of the same resolution [3].”
BackgroundMethodCloud-Based Networked Visual Servo Control
Wu, H., Lou, L., Chen, C.-C., Hirche, S., & Kuhnlenz, K. (2013) · IEEE Transactions on Industrial Electronics · doi:10.1109/tie.2012.2186775 · Cited by 92
“For example, it takes about 20ms to extract SURF (Speed Up Robust Features) [17] from an image (640 × 480pixels) on graphic card NVIDIA-GeForce 8800, while the sampling interval of the joint position control loop is typically in the range of 1ms.”
MethodInfluentialLi, K., Yao, J., Lu, X., Li, L., & Zhang, Z. (2016) · Neurocomputing · doi:10.1016/j.neucom.2015.07.137 · Cited by 95
Domain Adaptive Neural Networks for Object RecognitionPreprint
Ghifary, M., Kleijn, W. B., & Zhang, M. (2014) · arXiv (Cornell University) · doi:10.48550/arxiv.1409.6041 · Cited by 93
Palm vein recognition based on multi-sampling and feature-level fusion
Yan, X., Kang, W., Deng, F., & Wu, Q. (2014) · Neurocomputing · doi:10.1016/j.neucom.2014.10.019 · Cited by 92
Visual localization by linear combination of image descriptors
Torii, A., Šivic, J., & Pajdla, T. (2011) · IEEE International Conference on Computer Vision Workshops · doi:10.1109/iccvw.2011.6130230 · Cited by 95
Zaffar, M., Ehsan, S., Milford, M. J., & McDonald-Maier, K. D. (2020) · IEEE Robotics and Automation Letters · doi:10.1109/lra.2020.2969917 · Cited by 116
“Scale Invariant Feature Transform (SIFT [10]) and Speeded Up Robust Features (SURF [11]) are one of the most widely used local descriptors and have been applied to VPR problem in [12], [13], [14], [15], [16].”
MethodClassification of non-small cell lung cancer using one-dimensional convolutional neural network
Moitra, D., & Mandal, R. K. (2020) · Expert Systems with Applications · doi:10.1016/j.eswa.2020.113564 · Cited by 104
A Novel Hybrid Approach Based on Deep CNN to Detect Glaucoma Using Fundus Imaging
Mahum, R., Rehman, S. U., Okon, O. D., Alabrah, A. A., Meraj, T., & Rauf, H. T. (2021) · Electronics · doi:10.3390/electronics11010026 · Cited by 99
Object recognition datasets and challenges: A review
Salari, A., Djavadifar, A., Liu, X., & Najjaran, H. (2022) · Neurocomputing · doi:10.1016/j.neucom.2022.01.022 · Cited by 110
OrienterNet: Visual Localization in 2D Public Maps with Neural Matching
Sarlin, P.-E., DeTone, D., Yang, T.-Y., Avetisyan, A., Straub, J., Malisiewicz, T., Bulò, S. R., Newcombe, R. A., Kontschieder, P., & Balntas, V. (2023) · 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/cvpr52729.2023.02072 · Cited by 170
“These are composed of sparse point clouds, which are commonly built with Structure-fromMotion (SfM) [2, 22, 36, 43, 61, 67] from sparse points matched across multiple views [9, 38, 53].”
BackgroundGrid-guided Neural Radiance Fields for Large Urban Scenes
Xu, L., Xiangli, Y., Peng, S., Pan, X., Zhao, N., Theobalt, C., Dai, B., & Lin, D. (2023) · 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/cvpr52729.2023.00802 · Cited by 180
Computer Vision Onboard UAVs for Civilian Tasks
Campoy, P., Correa, J. F., Mondragon, I. F., Martínez, C., Olivares, M., Mejías, L., & Artieda, J. (2008) · Journal of Intelligent & Robotic Systems · doi:10.1007/s10846-008-9256-z · Cited by 116
“The SLAM algorithm is based on the features or corners’ matching process using SURF features [30] or on the Harris Corner detector [18].”
MethodA dataset and evaluation methodology for template-based tracking algorithms
Lieberknecht, S., Benhimane, S., Meier, P. G., & Navab, N. (2009) · Proceedings - International Symposium on Mixed and Augmented Reality, ISMAR · doi:10.1109/ismar.2009.5336487 · Cited by 103
“However, it should be mentioned that the evaluation of SIFT took more than 2.5 days (approximately 3 s per frame) to compute whereas FERNS, SURF and ESM finished in less than 6 hours each.”
MethodInfluentialX-RHex: A Highly Mobile Hexapedal Robot for Sensorimotor Tasks
Galloway, K. C., Haynes, G. C., Ilhan, B. D., Johnson, A. M., Knopf, R., Lynch, G. A., Plotnick, B. N., White, M., & Koditschek, D. E. (2010) · ScholarlyCommons (University of Pennsylvania) · Cited by 94
Scalar quantization for large scale image search
Zhou, W., Lu, Y., Li, H., & Tian, Q. (2012) · ACM International Conference on Multimedia (ACM MM) · doi:10.1145/2393347.2393377 · Cited by 97
“As a variation, SURF [9] demonstrates good performance but achieves better efficiency.”
BackgroundAdAttester: Secure Online Mobile Advertisement Attestation Using TrustZone
Li, W., Li, H., Chen, H., & Xia, Y. (2015) · Proceedings of the 13th Annual International Conference on Mobile Systems, Applications, and Services · doi:10.1145/2742647.2742676 · Cited by 91
“Example include SIFT [29] and SURF [12], which can tolerate complex transformations like rotation and scale down and have also recently been parallelized [15, 20].”
MethodLittleYOLO-SPP: A delicate real-time vehicle detection algorithm
S, S. J., & P, E. R. (2021) · Optik · doi:10.1016/j.ijleo.2020.165818 · Cited by 91
Aitken, J. M., Evans, M. H., Worley, R., Edwards, S., Zhang, R., Dodd, T. J., Mihaylova, L. S., & Anderson, S. (2021) · IEEE Access · doi:10.1109/access.2021.3115981 · Cited by 116
Interactive e-learning system using pattern recognition and augmented reality
Lee, S. U., Choi, J., & Park, J.-I. (2009) · IEEE Transactions on Consumer Electronics · doi:10.1109/tce.2009.5174470 · Cited by 95
Multi-modal authentication system for smartphones using face, iris and periocular
Raja, K. B., Ramachandra, R., Stokkenes, M., & Busch, C. (2015) · International Conference on Biometrics · doi:10.1109/icb.2015.7139044 · Cited by 103
“As compared to SIFT features, SURF descriptors are more robust to rotation, scale, illumination and contrast changes [2].”
ResultHierarchical Place Recognition for Topological Mapping
Garcia‐Fidalgo, E., & Ortiz, A. (2017) · IEEE Transactions on Robotics · doi:10.1109/tro.2017.2704598 · Cited by 93
Large-scale, real-time visual–inertial localization revisited
Lynen, S., Zeisl, B., Aiger, D., Bosse, M. C., Hesch, J. A., Pollefeys, M., Siegwart, R., & Sattler, T. (2020) · The International Journal of Robotics Research · doi:10.1177/0278364920931151 · Cited by 114
“Descriptors such as SIFT (Lowe 2004) and SURF (Bay et al. 2006) provide better performance under these conditions as well as in large scenes with strong visual aliasing.”
Methodİlhan, H. O., Sığırcı, I. O., Serbes, G., & Aydın, N. (2020) · Medical & Biological Engineering & Computing · doi:10.1007/s11517-019-02101-y · Cited by 105
“The SURF-algorithm [12] is based on a very similar principle with the Scale-Invariant Feature Transform (SIFT), which was proposed by Lowe [47].”
MethodSVT-Net: Super Light-Weight Sparse Voxel Transformer for Large Scale Place Recognition
Fan, Z., Song, Z., Liu, H., Lu, Z., He, J., & Du, X. (2022) · Proceedings of the AAAI Conference on Artificial Intelligence · doi:10.1609/aaai.v36i1.19934 · Cited by 119
Automatic targetless LiDAR–camera calibration: a survey
Li, X., Xiao, Y., Wang, B., Ren, H., Zhang, Y., & Ji, J. (2022) · Artificial Intelligence Review · doi:10.1007/s10462-022-10317-y · Cited by 119
RING++: Roto-Translation Invariant Gram for Global Localization on a Sparse Scan Map
Xu, X., Lu, S., Wu, J., Lu, H., Zhu, Q., Liao, Y., Xiong, R., & Wang, Y. (2023) · IEEE Transactions on Robotics · doi:10.1109/tro.2023.3303035 · Cited by 122
Classification of dermoscopy patterns using deep convolutional neural networks
Demyanov, S., Chakravorty, R., Abedini, M., Halpern, A., & Garnavi, R. (2016) · 2016 IEEE 13th International Symposium on Biomedical Imaging (ISBI) · doi:10.1109/isbi.2016.7493284 · Cited by 90
“We suppose it is because SIFT and SURF are not effective on small images with non discriminative key points.”
MethodInfluentialCross-View Action Recognition via Transferable Dictionary Learning
Zheng, J., Jiang, Z., & Chellappa, R. (2016) · IEEE Transactions on Image Processing · doi:10.1109/tip.2016.2548242 · Cited by 92
ICIP 2016 competition on mobile ocular biometric recognition
Rattani, A., Derakhshani, R. R., Saripalle, S. K., et al. (2016) · Proceedings - International Conference on Image Processing · doi:10.1109/icip.2016.7532371 · Cited by 89
“…is based on the weighted combination of four local descriptors for encoding local image textures, namely: Multiscale Local Binary Patterns (MLBP) [15], Phase Histogram of Oriented Gradients (PHOG) [16], Scale Invariant Feature Transform (SIFT) [17], and Speeded Up Robust Features (SURF) [14].”
BackgroundMethodDistribution-matching embedding for visual domain adaptation
Baktashmotlagh, M., Harandi, M. T., & Salzmann, M. (2016) · Journal of Machine Learning Research · doi:10.5555/2946645.3007061 · Cited by 109
Multi-camera visual SLAM for autonomous navigation of micro aerial vehicles
Yang, S., Scherer, S., Yi, X., & Zell, A. (2017) · Robotics and Autonomous Systems · doi:10.1016/j.robot.2017.03.018 · Cited by 101
“Compared with some other local feature descriptors, e.g. SIFT (Lowe, 2004), SURF (Bay et al., 2006), the BRIEF descriptor is not scale invariant nor rotation invariant.”
MethodJabeen, S., Mehmood, Z., Mahmood, T., Saba, T., Rehman, A. U., & Mahmood, M. T. (2018) · PLoS ONE · doi:10.1371/journal.pone.0194526 · Cited by 94
Advances in Video Compression System Using Deep Neural Network: A Review and Case Studies
Ding, D., Ma, Z., Chen, D., Chen, Q., Liu, Z. S.-J., & Zhu, F. M. (2021) · Proceedings of the IEEE · doi:10.1109/jproc.2021.3059994 · Cited by 116
“…was then extended using handcrafted local features, e.g. , the scale invariant feature transform (SIFT) [91], speeded up robust features (SURF) [92], and local binary patterns (LBP) [93] Recently, stacked DNNs have demonstrated their superior efficiency in many computer vision tasks, This…”
DFM: A Performance Baseline for Deep Feature Matching
Efe, U., İnce, K. G., & Alatan, A. A. (2021) · IEEE/CVF Conference on Computer Vision and Pattern Recognition Workshops (CVPRW) · doi:10.1109/cvprw53098.2021.00484 · Cited by 109
“The most popular algorithms [33, 6, 52, 30, 1, 2] work on both Feature Detection and Description.”
A deep analysis on age estimation
Huerta, I., Fernández, C., Segura, C., Hernando, J., & Prati, A. (2015) · Pattern Recognition Letters · doi:10.1016/j.patrec.2015.06.006 · Cited by 92
Heterogeneous Computing with OpenCL 2.0
(2015) · Elsevier eBooks · doi:10.1016/c2013-0-15490-6 · Cited by 89
Learning Image Descriptors with the Boosting-Trick
Trzcinski, T. P., Christoudias, C. M., Lepetit, V., & Fua, P. (2012) · Infoscience (Ecole Polytechnique Fédérale de Lausanne) · Cited by 115
Efficient processing of MRFs for unconstrained-pose face recognition
Arashloo, S. R., & Kittler, J. (2013) · IEEE International Conference on Biometrics: Theory, Applications and Systems (BTAS) · doi:10.1109/btas.2013.6712721 · Cited by 94
Karakasis, E. G., Amanatiadis, A. A., Gasteratos, A. C., & Chatzichristofis, S. A. (2015) · Pattern Recognition Letters · doi:10.1016/j.patrec.2015.01.005 · Cited by 92
Scalable Mobile Image Retrieval by Exploring Contextual Saliency
Yang, X., Qian, X., & Xue, Y. (2015) · IEEE Transactions on Image Processing · doi:10.1109/tip.2015.2411433 · Cited by 87
Receptive Fields Selection for Binary Feature Description
Fan, B., Kong, Q., Trzcinski, T. P., Wang, Z., Pan, C., & Fua, P. (2014) · IEEE Transactions on Image Processing · doi:10.1109/tip.2014.2317981 · Cited by 89
Markerless Vision-Based Augmented Reality for Urban Planning
Carozza, L., Tingdahl, D., Bosché, F., & Gool, L. V. (2012) · Computer-Aided Civil and Infrastructure Engineering · doi:10.1111/j.1467-8667.2012.00798.x · Cited by 100
Self-citationLeveraging deep neural networks to fight child pornography in the age of social media
Vitorino, P. R. R., Avila, S., Pérez, M., & Rocha, A. D. R. (2017) · Journal of Visual Communication and Image Representation · doi:10.1016/j.jvcir.2017.12.005 · Cited by 91
MARVEL: Enabling Mobile Augmented Reality with Low Energy and Low Latency
Chen, K., Li, T., Kim, H., Culler, D., & Katz, R. H. (2018) · ACM Conference on Embedded Networked Sensor Systems (SenSys) · doi:10.1145/3274783.3274834 · Cited by 106
“Then the SfM tool computes the relative 6-DOF location between every two consecutive images in the video stream (of a room) by matching their salient features (e.g., SURF [6]) and projects all the 2D feature points into a 3D coordinate system (a.k.a. frame), denoted as the model frame M in Figure 3.”
BackgroundMethodA Performance Evaluation of Local Features for Image-Based 3D Reconstruction
Fan, B., Kong, Q., Wang, X., Wang, Z., Xiang, S., Pan, C., & Fua, P. (2019) · IEEE Transactions on Image Processing · doi:10.1109/tip.2019.2909640 · Cited by 97
Multilayer pLSA for multimodal image retrieval
Lienhart, R. W., Romberg, S., & Hörster, E. (2009) · ACM International Conference on Image and Video Retrieval (CIVR) · doi:10.1145/1646396.1646408 · Cited by 88
“, SIFT [19], SURF [11], Geometric blur [2], or self-similarity feature [22]) using different sampling strategies (e.”
MethodAutonomous Visual Mapping and Exploration With a Micro Aerial Vehicle
Heng, L., Honegger, D., Lee, G. H., Meier, L., Tanskanen, P., Fraundorfer, F., & Pollefeys, M. (2014) · Journal of Field Robotics · doi:10.1002/rob.21520 · Cited by 85
“We extract the SURF features and descriptors (Bay et al., 2008) for every image and quantize them into visual words using a vocabulary tree which was pre-trained on a general image data set.”
MethodA Novel Neural Network for Remote Sensing Image Matching
Zhu, H., Jiao, L., Ma, W., Liu, F., & Zhao, W. (2019) · IEEE Transactions on Neural Networks and Learning Systems · doi:10.1109/tnnls.2018.2888757 · Cited by 88
Subjective and Objective Quality Assessment of Stitched Images for Virtual Reality
Madhusudana, P. C., & Soundararajan, R. (2019) · IEEE Transactions on Image Processing · doi:10.1109/tip.2019.2921858 · Cited by 98
Video Activity Recognition: State-of-the-Art
Rodríguez-Moreno, I., Martínez‐Otzeta, J. M., Sierra, B., Rodríguez, I., & Jauregi, E. (2019) · Sensors · doi:10.3390/s19143160 · Cited by 104
A Review of Deep Learning-based Human Activity Recognition on Benchmark Video Datasets
Sharma, V., Gupta, M., Pandey, A. K., Mishra, D., & Kumar, A. (2022) · Applied Artificial Intelligence · doi:10.1080/08839514.2022.2093705 · Cited by 113
Bilateral Functions for Global Motion Modeling
Lin, W.-Y., Cheng, M., Lu, J., Yang, H., Do, M. N., & Torr, P. H. S. (2014) · Lecture notes in computer science · doi:10.1007/978-3-319-10593-2_23 · Cited by 97
A mobile application of American sign language translation via image processing algorithms
Jin, C. M., Omar, Z., & Jaward, M. H. (2016) · IEEE Region 10 Symposium (TENSYMP) · doi:10.1109/tenconspring.2016.7519386 · Cited by 87
“The Hessian matrix with point and scale is defined as in (2): (2) where is the convolution of the image with the second-order derivative of the Gaussian as described by Bay et al. [11].”
MethodResultInfluentialAn Analysis and Implementation of the Harris Corner Detector
Sánchez, J., Monzón, N., & Salgado, A. (2018) · Image Processing On Line · doi:10.5201/ipol.2018.229 · Cited by 99
Hough Pyramid Matching: Speeded-Up Geometry Re-ranking for Large Scale Image Retrieval
Avrithis, Y., & Tolias, G. (2013) · International Journal of Computer Vision · doi:10.1007/s11263-013-0659-3 · Cited by 85
“We extract SURF features and descriptors [4] from each image, setting strength threshold to 2.”
MethodDetection of moving objects with a moving camera using non-panoramic background model
Kim, S. W., Kim, S. W., Yun, K., et al. (2012) · Machine Vision and Applications · doi:10.1007/s00138-012-0448-y · Cited by 86
Learning to Match Aerial Images with Deep Attentive Architectures
Altwaijry, H., Trulls, E., Hays, J., Fua, P., & Belongie, S. (2016) · 2016 IEEE Conference on Computer Vision and Pattern Recognition (CVPR) · doi:10.1109/cvpr.2016.385 · Cited by 84
Tobiasz, A., Markiewicz, J. S., Łapiński, S., Nikel, J., Kot, P., & Muradov, M. (2019) · Sustainability · doi:10.3390/su11247046 · Cited by 94
TIB-Net: Drone Detection Network With Tiny Iterative Backbone
Sun, H., Yang, J. J., Shen, J., Liang, D., Ning-Zhong, L., & Zhou, H. (2020) · IEEE Access · doi:10.1109/access.2020.3009518 · Cited by 103
“In previous studies, most object detection methods like HOG [2], SIFT [3], SURF [4], et al. extract handcrafted features for object detection.”
MethodLong short-term memory-based grade monitoring in froth flotation using a froth video sequence
Zhang, H., Tang, Z., Xie, Y., Gao, X., Chen, Q., & Gui, W. (2020) · Minerals Engineering · doi:10.1016/j.mineng.2020.106677 · Cited by 86
PDC-Net+: Enhanced Probabilistic Dense Correspondence Network
Truong, P., Danelljan, M., Timofte, R., & Gool, L. V. (2023) · IEEE Transactions on Pattern Analysis and Machine Intelligence · doi:10.1109/tpami.2023.3249225 · Cited by 146
“Keypoint detectors and descriptors are either handcrafted [10], [11], [12], [13] or learned [2], [14], [15], [16], [17], [18], [19].”
BackgroundSelf-citationXie, L., Wang, J., Zhang, B., & Tian, Q. (2015) · IEEE Transactions on Multimedia · doi:10.1109/tmm.2015.2408566 · Cited by 87
Robust Place Recognition With Stereo Sequences
Cadena, C., Galvez-López, D., Tardos, J. D., & Neira, J. (2012) · IEEE Transactions on Robotics · doi:10.1109/tro.2012.2189497 · Cited by 85
Enhancing Automatic Maritime Surveillance Systems With Visual Information
Bloisi, D. D., Previtali, F., Pennisi, A., Nardi, D., & Fiorini, M. (2016) · IEEE Transactions on Intelligent Transportation Systems · doi:10.1109/tits.2016.2591321 · Cited by 87
Optimal Bayesian Transfer Learning
Karbalayghareh, A., Qian, X., & Dougherty, E. R. (2018) · IEEE Transactions on Signal Processing · doi:10.1109/tsp.2018.2839583 · Cited by 101
Frustratingly Easy NBNN Domain Adaptation
Tommasi, T., & Barbara Caputo, B. (2013) · IEEE/CVF International Conference on Computer Vision (ICCV) · doi:10.1109/iccv.2013.116 · Cited by 78
Comparison of OpenCV's feature detectors and feature matchers
Noble, F. K. (2016) · International Conference on Mechatronics and Machine Vision in Practice (M2VIP) · doi:10.1109/m2vip.2016.7827292 · Cited by 105
“Given an application, the decision to use a specific feature detector or feature matcher may be based on a single objective, e.g. detect the greatest number of features or match features in Algorithm 5 input: enum matchFlag m if m == BF then mFlag == m if dFlag == SIFT ∣∣ SURF ∣∣ KAZE then type = float dist = NORM_L2…”
BackgroundInfluentialUpgrading Optical Flow to 3D Scene Flow Through Optical Expansion
Yang, G., & Ramanan, D. (2020) · 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/cvpr42600.2020.00141 · Cited by 98
“The concept of scale changes of visual features is well studied in the context of feature descriptor and matching [4, 31, 42] as well as dense optical flow [44, 57, 59].”
Background
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.