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 1,601–1,700 of 14,752
Sparse Color Interest Points for Image Retrieval and Object Categorization
Stöttinger, J., Hanbury, A., Sebe, N., & Gevers, T. (2012) · IEEE Transactions on Image Processing · doi:10.1109/tip.2012.2186143 · Cited by 47
Markerless Endoscopic Registration and Referencing
Wengert, C., Cattin, P. C., Duff, J. M., et al. (2006) · Lecture notes in computer science · doi:10.1007/11866565_100 · Cited by 45
“From these two views interest points and their region descriptors are extracted using the method described in [17].”
MethodRobust and effective component-based banknote recognition by SURF features
Hasanuzzaman, F. M., Yang, X., & Tian, Y. (2011) · Wireless and Optical Communications Conference · doi:10.1109/wocc.2011.5872294 · Cited by 45
Content-Based Image Retrieval using Local Features Descriptors and Bag-of-Visual Words
Alkhawlani, M., Elmogy, M. M., & El-Bakry, H. M. (2015) · International Journal of Advanced Computer Science and Applications · doi:10.14569/ijacsa.2015.060929 · Cited by 45
“The system tries to bring more accuracy with the option to use the two main local descriptors (SIFT [6], SURF [7]).”
BackgroundMethodMethodology for partial fingerprint enrollment and authentication on mobile devices
Mathur, S., Vjay, A., Shah, J., et al. (2016) · International Conference on Biometrics · doi:10.1109/icb.2016.7550093 · Cited by 45
“It is proven that repeatability and distinctiveness of KAZE features is better than SIFT and SURF, but is computationally very expensive due to use of Additive Operator Splitting (AOS) schemes to solve nonlinear diffusion equations.”
MethodACACIA: Context-aware Edge Computing for Continuous Interactive Applications over Mobile Networks
Cho, J., Sundaresan, K., Mahindra, R., der Merwe, J. V., & Rangarajan, S. (2016) · ACM International Conference on emerging Networking EXperiments and Technologies (CoNEXT) · doi:10.1145/2999572.2999604 · Cited by 45
A Real Time Object Recognition and Counting System for Smart Industrial Camera Sensor
Lee, S., & Yang, C. (2017) · IEEE Sensors Journal · doi:10.1109/jsen.2017.2671457 · Cited by 45
“point) [9], ORB (Oriented FAST and Rotated BRIEF) [10], SIFT (Scale-Invarient Feature Transform) [11], and SURF (Speeded Up Robust Feature) [12], etc.”
Fusion-based underwater image enhancement by wavelet decomposition
Wang, Y., Ding, X., Wang, R., Zhang, J., & Fu, X. (2017) · IEEE International Conference on Industrial Technology (ICIT) · doi:10.1109/icit.2017.7915500 · Cited by 52
“Moreover, the SURF operator is also applied to the enhanced images that corresponding to those original degraded underwater images.”
MethodInfluentialContext-aware obstacle detection for navigation by visually impaired
Gharani, P., & Karimi, H. A. (2017) · Image and Vision Computing · doi:10.1016/j.imavis.2017.06.002 · Cited by 52
Fast reflective offset-guided searching method for copy-move forgery detection
Bi, X., & Pun, C. (2017) · Information Sciences · doi:10.1016/j.ins.2017.08.044 · Cited by 54
“SIFT[2] SURF[2] Bravo[22] SBFD[11] ASBFD[10] DFFD[6] Proposed Method (f) 100 90 80 70 60 50 40 30 20 0 0.1 0.2 0.3 0.4 0.5 0.6 0.7 0.8 0.9 1 P re ci si on JPEG-compression[quality-factor] SIFT[2] SURF[2] Bravo[22] SBFD[11] ASBFD[10] DFFD[6] Proposed Method (g) 100 90 80 70 60 50 40 30 20 0 0.1 0.2 0.3 0.4 0.5 0.6 0.7…”
MethodInfluentialDeng, W., Hu, J., & Guo, J. H. (2018) · IEEE Transactions on Pattern Analysis and Machine Intelligence · doi:10.1109/tpami.2018.2800008 · Cited by 49
“For example, regions of interest are typically represented by handcrafted SIFT [3], SURF [9] and BRIEF [10] descriptors and their variants.”
MethodMapping Forest Structure Using UAS inside Flight Capabilities
Kuželka, K., & Surový, P. (2018) · Sensors · doi:10.3390/s18072245 · Cited by 51
“This approach, which simply requires a commercial off-the-shelf digital camera, is a recent development triggered by advances in computer vision, such as the scale-invariant feature transform (SIFT) [14] and speeded up robust features (SURF) [15] algorithms.”
MethodLydon, D., Lydon, M., del Rincón, J. M., et al. (2018) · IEEE Sensors Journal · doi:10.1109/jsen.2018.2853646 · Cited by 45
Collision Avoidance for Redundant Robots in Position-Based Visual Servoing
Shi, H., Chen, J., Pan, W., et al. (2018) · IEEE Systems Journal · doi:10.1109/jsyst.2018.2865503 · Cited by 45
“A variety of efficient methods are used for feature extraction and matching, such as RGB-based methods, SIFT-based methods, SURF-based methods [9], [10], and subfeature selection methods [11]–[14].”
MethodLSTM Accelerator for Convolutional Object Identification
Savvopoulos, A., Kanavos, A., Phivos Mylonas, B., & Sioutas, S. (2018) · Algorithms · doi:10.3390/a11100157 · Cited by 47
CURE-OR: Challenging Unreal and Real Environments for Object Recognition
Temel, D., Lee, J., & AlRegib, G. (2018) · International Conference on Machine Learning and Applications · doi:10.1109/icmla.2018.00028 · Cited by 56
“Traditionally, these state-of-the-art recognition approaches were based on handcrafted features such as SIFT [3], SURF [4] and HoG [5].”
MethodDLD: A Deep Learning Based Line Descriptor for Line Feature Matching
Lange, M., Schweinfurth, F., & Schilling, A. (2019) · IEEE/RSJ International Conference on Intelligent Robots and Systems (IROS) · doi:10.1109/iros40897.2019.8968062 · Cited by 52
Adaptive path finding algorithm in dynamic environment for warehouse robot
Ng, M.-K., Chong, Y.-W., Ko, K., Park, Y.-H., & Leau, Y. (2020) · Neural Computing and Applications · doi:10.1007/s00521-020-04764-3 · Cited by 57
“SURF is a patented local feature detector and descriptor that used for object recognition [19].”
BackgroundXu, Z., Liu, Y., Shi, X., Wang, Y., & Zheng, Y. (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/cvpr42600.2020.00602 · Cited by 51
“[6] presents an efficient technique to approximate the first and second order Gaussian derivatives by using box filters and integral images, which is a contributory factor for the efficiency of the benchmark detector and descriptor.”
MethodZou, X. J., & Pan, B. (2020) · Optics and Lasers in Engineering · doi:10.1016/j.optlaseng.2020.106432 · Cited by 45
An Asynchronous Kalman Filter for Hybrid Event Cameras
Wang, Z., Ng, Y., Scheerlinck, C., & Mahony, R. E. (2021) · IEEE/CVF International Conference on Computer Vision (ICCV) · doi:10.1109/iccv48922.2021.00050 · Cited by 62
Karaman, S., Benois‐Pineau, J., Dovgalecs, V., Pinquier, J., Andre-Obrecht, R., Gaëstel, Y., & Dartigues, F. (2014) · Cited by 45
Cooper, N., Bernal-Cárdenas, C., Chaparro, O., Moran, K., & Poshyvanyk, D. (2021) · ACM/IEEE International Conference on Software Engineering: Companion Proceedings (ICSE-Companion) · doi:10.1109/icse43902.2021.00091 · Cited by 81
Approximate Gaussian Mixtures for Large Scale Vocabularies
Avrithis, Y., & Kalantidis, Y. (2012) · Lecture notes in computer science · doi:10.1007/978-3-642-33712-3_2 · Cited by 50
“We extract SURF [30] features and descriptors from each image.”
MethodReal-time mobile recipe recommendation system using food ingredient recognition
Maruyama, T., Kawano, Y., & Yanai, . (2012) · doi:10.1145/2390821.2390830 · Cited by 45
“To prevent information loss due to binarization of feature vectors, we use SURF local feature [2].”
MethodHuo, Y., Li, Z., & Zhang, F. (2020) · IEEE Access · doi:10.1109/access.2020.3041415 · Cited by 44
“The features and descriptors usually produced by handcrafted detectors, e.g., SIFT [23], SURF [24], BRIEF [25], Canny edge detector [26], and Hough Transform [27], etc.”
MethodDynamic Concept Composition for Zero-Example Event Detection
Chang, X., Yang, Y., Long, G., Zhang, C., & Hauptmann, A. G. (2016) · Proceedings of the AAAI Conference on Artificial Intelligence · doi:10.1609/aaai.v30i1.10474 · Cited by 45
Snapshot hyperspectral imaging polarimetry with full spectropolarimetric resolution
Mu, T., Han, F., Li, H., et al. (2021) · Optics and Lasers in Engineering · doi:10.1016/j.optlaseng.2021.106767 · Cited by 44
GIFdroid: automated replay of visual bug reports for Android apps
Feng, S., & Chen, C. (2022) · ACM/IEEE International Conference on Software Engineering: Companion Proceedings (ICSE-Companion) · doi:10.1145/3510003.3510048 · Cited by 67
“Baselines: To further demonstrate the advantage of our method, we compare it with 10 image processing baselines, including pixel level (e.g, euclidean distance [33], color histogram [72], fingerprint [16]), and structural level (e.g., SSIM [73], SIFT [52], SURF [21], ORB [63]).”
Tsourounis, D., Kastaniotis, D., Theoharatos, C., Kazantzidis, A., & Economou, G. (2022) · Journal of Imaging · doi:10.3390/jimaging8100256 · Cited by 51
“Hand-crafted features have been extensively used in computer vision problems, mainly for the task of image classification [1–3].”
MethodWaterNeRF: Neural Radiance Fields for Underwater Scenes
Sethuraman, A. V., Ramanagopal, M. S., & Skinner, K. A. (2023) · MTS/IEEE OCEANS Conference (OCEANS) · doi:10.23919/oceans52994.2023.10336972 · Cited by 82
A survey of feature matching methods
Huang, Q., Guo, X., Wang, Y., Sun, H., & Yang, L. (2024) · IET Image Processing · doi:10.1049/ipr2.13032 · Cited by 70
BoQ: A Place is Worth a Bag of Learnable Queries
Ali-bey, A., Chaib-draa, B., & Giguère, P. (2024) · 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/cvpr52733.2024.01685 · Cited by 116
“Early visual place recognition (VPR) techniques relied on hand-crafted local features such as SIFT [33], SURF [9] and ORB [41], which were aggregated into global descriptors using techniques like Bag-of-Words (BoW) [20, 39, 45] or Vector of Locally Aggregated Descriptors (VLAD) [5, 27].”
MethodLearning to segment dense cell nuclei with shape prior
Lou, X., Köthe, U., Wittbrodt, J., & Hamprecht, F. A. (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.2012.6247778 · Cited by 49
Fast and robust object segmentation with the Integral Linear Classifier
Aldavert, D., de Mántaras, R. L., Arnau Ramisa, A., & Toledo, R. (2010) · 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.2010.5540098 · Cited by 42
Surf feature extraction in encrypted domain
Bai, Y., Li, Z., Cheng, B. H., et al. (2014) · IEEE International Conference on Multimedia and Expo (ICME) · doi:10.1109/icme.2014.6890170 · Cited by 44
Visual SLAM and Moving-object Detection for a Small-size Humanoid Robot
Wang, Y.-T., Lin, M.-C., & Ju, R.-C. (2010) · International Journal of Advanced Robotic Systems · doi:10.5772/9700 · Cited by 47
Efficient object detection using cascades of nearest convex model classifiers
Çevıkalp, H., & Triggs, B. (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.2012.6248047 · Cited by 46
“Many of these are based on oriented image gradients, including SIFT [16], SURF [4], Histogram of Oriented Gradients (HOG) [6], PHOG [26], Generalized Shape Con-text [5] and Local Edge Orientation Histograms [15].”
MethodRANSAC for motion-distorted 3D visual sensors
Anderson, S., & Barfoot, T. D. (2013) · IEEE/RSJ International Conference on Intelligent Robots and Systems (IROS) · doi:10.1109/iros.2013.6696649 · Cited by 48
Autonomous Localization in Unknown Environments
Callmer, J. (2013) · DiVA (Linkoping University) · Cited by 48
Probabilistic Model for Robust Affine and Non-Rigid Point Set Matching
Qu, H., Wang, J., Li, B., & Yu, M. (2016) · IEEE Transactions on Pattern Analysis and Machine Intelligence · doi:10.1109/tpami.2016.2545659 · Cited by 47
“One solution is that the correspondence is first established by these feature extractors, such as SIFT [12] or SURF [13].”
MethodBlock-based adaptive ROI for remote photoplethysmography
Po, L.-M., Feng, L., Li, Y., Xu, X., Cheung, T. C.-H., & Cheung, K.-W. (2017) · Multimedia Tools and Applications · doi:10.1007/s11042-017-4563-7 · Cited by 46
SuperPatchMatch: An Algorithm for Robust Correspondences Using Superpixel Patches
Giraud, R., Ta, V., Bugeau, A., Coupé, P., & Papadakis, N. (2017) · IEEE Transactions on Image Processing · doi:10.1109/tip.2017.2708504 · Cited by 43
“They have progressively proven their efficiency for several applications such as texture synthesis [36] and image denoising [37], and in the design of computer vision descriptors [38], [39] that include spatial information.”
Learning non-metric visual similarity for image retrieval
García, N., & Vogiatzis, G. (2019) · Image and Vision Computing · doi:10.1016/j.imavis.2019.01.001 · Cited by 48
“One common approach is to represent the local visual content of images by a set of hand-crafted features, such as SIFT [21], SURF [3] or BRIEF [5].”
BackgroundMethodNejad, E. M., Affendey, L. S., Latip, R. B., & Ishak, I. (2017) · International Conference on Imaging, Signal Processing and Communications (ICISPC) · doi:10.1145/3132300.3132331 · Cited by 53
“[11] Bay, H., Tuytelaars, T., Van Gool, L., Leonardis, A., Bischof, H., and Pinz, A. “SURF: Speeded Up Robust Features,” Comput.”
MethodHou, Y., Zhang, H., & Zhou, S. (2017) · Autonomous Robots · doi:10.1007/s10514-017-9684-3 · Cited by 43
“Local feature-based algorithms (Kosecka and Li 2004; Li and Kosecka 2006; Cummins and Newman 2011; Zhang 2011) use local features, such as SIFT (Lowe 2004) and SURF (Bay et al. 2006), to represent an image.”
MethodInfluentialLearning-Based Tone Mapping Operator for Efficient Image Matching
Rana, A., Valenzise, G., & Frederic Dufaux, A. (2018) · IEEE Transactions on Multimedia · doi:10.1109/tmm.2018.2839885 · Cited by 44
Drone-Based Vacant Parking Space Detection
Peng, C.-F., Hsieh, J.-W., Leu, S.-W., & Chuang, C.-H. (2018) · International Conference on Advanced Information Networking and Applications Workshops (WAINA) · doi:10.1109/waina.2018.00155 · Cited by 48
Vessel Optimal Transport for Automated Alignment of Retinal Fundus Images
Motta, D., Casaca, W. C. D. O., & Paiva, A. (2019) · IEEE Transactions on Image Processing · doi:10.1109/tip.2019.2925287 · Cited by 47
“Their approach relies on image gradients as the PIIFD descriptor combined with a robust local feature detector called Speeded-Up Robust Features (SURF) [43].”
MethodGuide to Match: Multi-Layer Feature Matching With a Hybrid Gaussian Mixture Model
Sun, K., Tao, W., & Qian, Y. (2019) · IEEE Transactions on Multimedia · doi:10.1109/tmm.2019.2957984 · Cited by 47
“Since SIFT [27] has been proposed about 20 years ago, a variety of feature descriptors [35]–[37] are already off-the-shelf for real application.”
BackgroundA Flexible and Efficient Real-Time ORB-Based Full-HD Image Feature Extraction Accelerator
Sun, R., Qian, J., Hung, J. L. R., Gong, Z., Miao, R., Xue, W., & Liu, P. (2019) · IEEE Transactions on Very Large Scale Integration (VLSI) Systems · doi:10.1109/tvlsi.2019.2945982 · Cited by 46
“The above architecture and dataflow can also be used for the other image feature extraction algorithms like SIFT and SURF.”
InfluentialFast and Incremental Loop Closure Detection Using Proximity Graphs
An, S., Che, G., Zhou, F., Liu, X., Ma, X., & Chen, Y. (2019) · IEEE/RSJ International Conference on Intelligent Robots and Systems (IROS) · doi:10.1109/iros40897.2019.8968043 · Cited by 51
“The methods above use local features such as SURF [13] and BRIEF [14].”
MethodInfluentialA Two-Stream Symmetric Network with Bidirectional Ensemble for Aerial Image Matching
Park, J.-H., Nam, W.-J., & Lee, S. (2020) · Remote Sensing · doi:10.3390/rs12030465 · Cited by 45
“Quantitative results Aerial Image Dataset Table 1 shows quantitative comparisons to the conventional computer vision methods (SURF [2], SIFT [1], ASIFT [4] + RANSAC [5] and OA-Match [61]) and CNNGeo [36] on aerial image data with large transformation.”
BackgroundMethodInfluentialFast Temporal Video Segmentation Based on Krawtchouk-Tchebichef Moments
Abdulhussain, S. H., Al-Haddad, S. A. R., Saripan, M. I., et al. (2020) · IEEE Access · doi:10.1109/access.2020.2987870 · Cited by 44
“Speeded up robust features (SURF) [32], scale-invariant feature transform (SIFT) [33], and Harris corner detector [34] are methods for LKP extraction.”
MethodVision‐based vehicle detection for road traffic congestion classification
Chetouane, A., Mabrouk, S., Jemili, I., & Mosbah, M. (2020) · Concurrency and Computation Practice and Experience · doi:10.1002/cpe.5983 · Cited by 49
“The second subcategory is Local Features Descriptors such as the Histograms Of Gradients (HOG), 24 Harr-like features, 25 Speed-Up Robust Features (SURF), 26 and Scale-Invariant Feature Transform (SIFT).”
BackgroundImbalanced Data Classification via Cooperative Interaction Between Classifier and Generator
Choi, H.-S., Jung, D., Kim, S., & Yoon, S. (2021) · IEEE Transactions on Neural Networks and Learning Systems · doi:10.1109/tnnls.2021.3052243 · Cited by 47
Moghimi, A., Sarmadian, A., Mohammadzadeh, A., et al. (2021) · IEEE Transactions on Geoscience and Remote Sensing · doi:10.1109/tgrs.2021.3063151 · Cited by 44
Jiang, S., Guo, C., Song, P., Wang, T., Wang, R., Zhang, T., Wu, Q., Pandey, R., & Zheng, G. (2022) · Lab on a Chip · doi:10.1039/d2lc00084a · Cited by 48
Interest Point Detectors for Visual SLAM
Mozos, Ó. M., Gil, A., Ballesta, M., & Reinoso, Ó. (2007) · Lecture notes in computer science · doi:10.1007/978-3-540-75271-4_18 · Cited by 45
PERFORMANCE EVALUATION OF A STRUCTURE AND MOTION STRATEGY IN ARCHITECTURE AND CULTURAL HERITAGE
Roncella, R., Re, C., & Forlani, G. (2012) · 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/isprsarchives-xxxviii-5-w16-285-2011 · Cited by 51
“Recent feature-based matching (Bay et al., 2008) and Structure from Motion (SFM) (Hartley and Zisserman, 2004) algorithms have indeed evolved by strengthening their mathematical, statistical and geometrical foundations in a fruitful competition that should improve their robustness and overall performance.”
BackgroundMethodHistory and Future of Tracking for Mobile Phone Augmented Reality
Wägner, D., & Schmalstieg, D. (2009) · International Symposium on Ubiquitous Virtual Reality · doi:10.1109/isuvr.2009.11 · Cited by 44
Fast and accurate near-duplicate image search with affinity propagation on the ImageWeb
Xie, L., Tian, Q., Zhou, W., & Zhang, B. (2014) · Computer Vision and Image Understanding · doi:10.1016/j.cviu.2013.12.011 · Cited by 42
“Popular descriptors for image retrieval include SIFT [2,10], SURF [11], BRIEF [12] and so on.”
MethodStructure from motion using dense CNN features with keypoint relocalization
Widya, A. R., Torii, A., & Okutomi, M. (2018) · IPSJ Transactions on Computer Vision and Applications · doi:10.1186/s41074-018-0042-y · Cited by 52
“In this decade, the performance of local features, namely, SIFT [11] and its variants [16, 21, 22, 23, 24] are validated on 3D reconstruction as well as many other tasks [25, 26, 27].”
BackgroundEnd-to-End Image Stitching Network via Multi-Homography Estimation
Song, D.-Y., Um, G., Lee, H. K., & Cho, D. (2021) · IEEE Signal Processing Letters · doi:10.1109/lsp.2021.3070525 · Cited by 43
Zhou, H., & Jayender, J. (2021) · Lecture notes in computer science · doi:10.1007/978-3-030-87202-1_32 · Cited by 62
Content-Aware Unsupervised Deep Homography Estimation and Its Extensions
Liu, S., Ye, N., Wang, C., Luo, K., Wang, J., & Sun, J. (2022) · IEEE Transactions on Pattern Analysis and Machine Intelligence · doi:10.1109/tpami.2022.3174130 · Cited by 58
“To solve a homography, traditional approaches often detect and match image features, such as classic SIFT [14], SURF [18], ORB [19] and Fig.”
MethodA Pedestrian Detection Network Model Based on Improved YOLOv5
Li, M.-L., Sun, G., & Yu, J.-X. (2023) · Entropy · doi:10.3390/e25020381 · Cited by 49
Multimodal Image Fusion Framework for End-to-End Remote Sensing Image Registration
Li, L., Han, L., Ding, M., & Cao, H. (2023) · IEEE Transactions on Geoscience and Remote Sensing · doi:10.1109/tgrs.2023.3247642 · Cited by 51
“Various feature-based methods have been developed over the past decades, represented by the scale-invariant feature transform (SIFT) [7], including these improved SIFT methods, such as SURF [8], PCA-SIFT [9], synthetic aperture radar (SAR)-SIFT [10], and DSP-SIFT [11].”
MethodEfficient feature selection using BoWs and SURF method for leaf disease identification
Bhagat, M., & Kumar, D. (2023) · Multimedia Tools and Applications · doi:10.1007/s11042-023-14625-5 · Cited by 48
“Integral images, as described by [2, 31], are briefly discussed in order to help the work stand on its own as a unit.”
BackgroundMethodResultInfluentialA survey of video-based human action recognition in team sports
Yin, H., Sinnott, R., & Jayaputera, G. (2024) · Artificial Intelligence Review · doi:10.1007/s10462-024-10934-9 · Cited by 76
A SURF-Based Spatio-Temporal Feature for Feature-Fusion-Based Action Recognition
Noguchi, A., & Yanai, . (2012) · Lecture notes in computer science · doi:10.1007/978-3-642-35749-7_12 · Cited by 45
“In the original method, we detect interest points and extract feature vectors employing the SURF method [7], and then we select moving interest points employing the Lucas-Kanade method [8].”
MethodInfluentialPalm Vein Recognition Based on Three Local Invariant Feature Extraction Algorithms
Pan, M., & Kang, W. (2011) · Lecture notes in computer science · doi:10.1007/978-3-642-25449-9_15 · Cited by 43
Review and classification of vision‐based localisation techniques in unknown environments
Afia, A. B., Deambrogio, L., Salós, D., Escher, A., Macabiau, C., Soulier, L., & Gay‐Bellile, V. (2014) · IET Radar Sonar & Navigation · doi:10.1049/iet-rsn.2013.0389 · Cited by 54
Automatic lesion detection in wireless capsule endoscopy — A simple solution for a complex problem
Iakovidis, D. K., & Koulaouzidis, A. (2014) · Proceedings - International Conference on Image Processing · doi:10.1109/icip.2014.7025453 · Cited by 57
Condition-invariant, top-down visual place recognition
Milford, M. J., Scheirer, W. J., Vig, E., Glover, A., Baumann, O., Mattingley, J. B., & Cox, D. (2014) · Proceedings - IEEE International Conference on Robotics and Automation/Proceedings · doi:10.1109/icra.2014.6907678 · Cited by 42
Online calibration of stereo rigs for long-term autonomy
Warren, M., McKinnon, D. H., & Upcroft, B. (2013) · Proceedings - IEEE International Conference on Robotics and Automation/Proceedings · doi:10.1109/icra.2013.6631096 · Cited by 43
“Initial pose is setup by first computing an essential matrix E 1 using 5 matched SURF [12] features using MLESAC [13] for robustness, from which the initial camera poses, P 0 and P(1) 0 , are extracted [14].”
MethodReview of Sensor-based Activity Recognition Systems
Guan, D., Yuan, W., Sarkar, A. J., Ma, T., & Lee, Y.-K. (2011) · IETE Technical Review · doi:10.4103/0256-4602.85975 · Cited by 54
Estimating Position of Mobile Robots From Omnidirectional Vision Using an Adaptive Algorithm
Li, L., Liu, Y.-H., Wang, K., & Fang, M. (2014) · IEEE Transactions on Cybernetics · doi:10.1109/tcyb.2014.2357797 · Cited by 45
“1) Feature Extracting and Tracking: In the experiment, we used the speeded up robust features (SURF) algorithm [36] to extract the image feature points and build up the feature descriptors.”
MethodScalable Object Retrieval with Compact Image Representation from Generic Object Regions
Sun, S., Zhou, W., Tian, Q., & Li, H. (2015) · ACM Transactions on Multimedia Computing Communications and Applications · doi:10.1145/2818708 · Cited by 46
“In addition, some works extend it with detectors such as the Hessian-Affine detector [Mikolajczyk and Schmid 2004], MSER detector [Kadir et al. 2004], and SURF [Bay et al. 2006].”
Method6D Relocalisation for RGBD Cameras Using Synthetic View Regression
Gee, A. P., & Mayol-Cuevas, W. W. (2012) · British Machine Vision Conference (BMVC) · doi:10.5244/c.26.113 · Cited by 41
SUSurE: Speeded Up Surround Extrema feature detector and descriptor for realtime applications
Ebrahimi, M., & Mayol-Cuevas, W. W. (2009) · 2009 IEEE Computer Society Conference on Computer Vision and Pattern Recognition Workshops · doi:10.1109/cvprw.2009.5204313 · Cited by 48
Semiautomated intraocular laser surgery using handheld instruments
Becker, B. C., MacLachlan, R. A., Lobes, L. A., & Riviere, C. N. (2010) · Lasers in Surgery and Medicine · doi:10.1002/lsm.20897 · Cited by 47
“Interest points can be matched based on similarity of SURF descriptors calculated at each interest point.”
Accurate and robust registration for in-hand modeling
Weise, T., Leibe, B., & Gool, L. V. (2008) · 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.2008.4587832 · Cited by 48
Self-citationState estimation of an underwater robot using visual and inertial information
Shkurti, F., Rekleitis, I. M., Scaccia, M., & Dudek, G. (2011) · IEEE/RSJ International Conference on Intelligent Robots and Systems (IROS) · doi:10.1109/iros.2011.6094680 · Cited by 45
An Object Tracking System Based on SIFT and SURF Feature Extraction Methods
Sakai, Y., Oda, T., Ikeda, M., & Barolli, L. (2015) · International Conference on Network-Based Information Systems · doi:10.1109/nbis.2015.121 · Cited by 44
6-DOF Image Localization From Massive Geo-Tagged Reference Images
Song, Y., Chen, X., Wang, X., et al. (2016) · IEEE Transactions on Multimedia · doi:10.1109/tmm.2016.2568743 · Cited by 43
“efficiency, SURF features [35] are extracted as the local features from the images in reference dataset.”
RepMatch: Robust Feature Matching and Pose for Reconstructing Modern Cities
Lin, W.-Y., Liu, S., Jiang, N., Do, M. N., Tan, P., & Lu, J. (2016) · Lecture notes in computer science · doi:10.1007/978-3-319-46448-0_34 · Cited by 41
Long-Term Time-Sensitive Costs for CRF-Based Tracking by Detection
Le, N., Heili, A., & Odobez, J. (2016) · Lecture notes in computer science · doi:10.1007/978-3-319-48881-3_4 · Cited by 43
“Although SURF matching can be discriminative for objects, it is less effective in human tracking because of clothing similarity, and data resolution where most features are found on human boundaries rather than within.”
MethodInfluentialLoop closure detection for visual SLAM using PCANet features
Xia, Y., Li, J., Qi, L., & Fan, H. (2016) · International Joint Conference on Neural Networks (IJCNN) · doi:10.1109/ijcnn.2016.7727481 · Cited by 44
Encoding the description of image sequences: A two-layered pipeline for loop closure detection
Bampis, L., Amanatiadis, A. A., & Gasteratos, A. C. (2016) · IEEE/RSJ International Conference on Intelligent Robots and Systems (IROS) · doi:10.1109/iros.2016.7759667 · Cited by 42
Convolutional bag of words for diabetic retinopathy detection from eye fundus images
Costa, P., & Campilho, A. (2017) · IPSJ Transactions on Computer Vision and Applications · doi:10.1186/s41074-017-0023-6 · Cited by 43
QuEst: A Quaternion-Based Approach for Camera Motion Estimation From Minimal Feature Points
Fathian, K., Ramirez-Paredes, J. P., Doucette, E. A., Curtis, J. W., & Gans, N. R. (2018) · IEEE Robotics and Automation Letters · doi:10.1109/lra.2018.2792142 · Cited by 56
“) that can be detected and matched in two or more images [7].”
BackgroundAn intelligent recommendation system using gaze and emotion detection
Jaiswal, S., Virmani, S., Sethi, V., De, K., & Roy, P. P. (2018) · Multimedia Tools and Applications · doi:10.1007/s11042-018-6755-1 · Cited by 46
“To obtain the reference point, initially an arbitrary “virtual reference point” is chosen and then keypoints that uniquely identify image are calculated using SURF algorithm [2] (Refer Fig.”
MethodReal-Time Wide-Baseline Place Recognition Using Depth Completion
Maffra, F., Teixeira, L., Chen, Z., & Chli, M. (2019) · IEEE Robotics and Automation Letters · doi:10.1109/lra.2019.2895826 · Cited by 47
New shape descriptor in the context of edge continuity
Susan, S., Agrawal, P., Mittal, M., & Bansal, S. (2019) · CAAI Transactions on Intelligence Technology · doi:10.1049/trit.2019.0002 · Cited by 46
“Signal Process., 2018, 29, (4), pp. 1269–1284 [13] Valgren, C., Lilienthal, A.J.: ‘SIFT, SURF & seasons: appearance-based long-term localization in outdoor environments’, Robot.”
ResultInfluentialYu, Y., Kurnianggoro, L., & Jo, K.-H. (2019) · International Journal of Control Automation and Systems · doi:10.1007/s12555-018-0234-3 · Cited by 43
“Some feature similarity methods such as BRIEF [26], SURF [27], and SIFT [28] can be used to instead of the optical flow from x t (cid:0) 1 to x t .”
MethodComputational Intelligence in Remote Sensing Image Registration: A survey
Wu, Y., Liu, J., Zhu, C.-Z., Bai, Z., Miao, Q., Ma, W., & Gong, M. (2020) · Machine Intelligence Research · doi:10.1007/s11633-020-1248-x · Cited by 47
Semi-Supervised Dim and Small Infrared Ship Detection Network Based on Haar Wavelet
Song, Z., Yang, J., Zhang, D., Wang, S., & Li, Z. (2021) · IEEE Access · doi:10.1109/access.2021.3058526 · Cited by 47
“Traditional detection methods, such as SIFT [1], HOG [2], SURF [3], et al. extract the features of the target through the handcrafted features, and put the extracted features into SVM [4] for classification and recognition.”
MethodRani, A., Jain, A. K., & Kumar, M. (2021) · Multimedia Tools and Applications · doi:10.1007/s11042-021-10810-6 · Cited by 43
“SURF is presented by Bay et al. [4] in 2006 and this algorithm is extended form of scale invariant feature transform (SIFT).”
MethodA 3D reconstruction method for buildings based on monocular vision
Xu, B., & Liu, C. (2021) · Computer-Aided Civil and Infrastructure Engineering · doi:10.1111/mice.12715 · Cited by 43
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.