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 10,601–10,700 of 14,752
Zhang, J., Zhang, J., & Wu, Z. (2025) · Structure and Infrastructure Engineering · doi:10.1080/15732479.2025.2532178 · Cited by 1
Zhao, Y., Fu, D.-J., Li, Z., & Feng, W. (2025) · Journal of Electrical Engineering and Technology · doi:10.1007/s42835-025-02358-4 · Cited by 1
Evaluating Data Collection Methods for Vision-Based Learning in Humanoid Robot Soccer
Hong, E. R., Ahn, J. S., Lee, S., Alvarez, A. F., Wang, S., & Hong, D. W. (2025) · International Conference on Ubiquitous Robots (UR) · doi:10.1109/ur65550.2025.11078062 · Cited by 1
“Traditionally, the vision task has been achieved using an amalgamation of classical & deep learning based approaching, including color segmentation, pattern detection [2] [3], feature extraction [9] [10] [11], and matching [12].”
Joint Optimization of Neural Radiance Fields and Continuous Camera Motion from a Monocular Video
Nguyen, H. C., Mao, W., Álvarez, J. M., et al. (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/cvpr52734.2025.01071 · Cited by 1
“However, this process introduces additional computation, lacks differentiability, and is sensitive to hand-crafted features [1, 17].”
Hand-Eye Coordinated Grasping Method for Textured Targets in Unstructured Dynamic Scenes
Luo, Y., Ruan, S., Li, Y., & Chen, D. (2025) · Chinese Journal of Mechanical Engineering · doi:10.1186/s10033-025-01344-1 · Cited by 1
Video Stabilization Algorithm Based on View Boundary Synthesis
Shan, W., Zhao, H., Li, X., et al. (2025) · Symmetry · doi:10.3390/sym17081351 · Cited by 1
Deep Learning Based Object Detection and its Application: A Review
Kumari, L., & Majumder, A. (2025) · SN Computer Science · doi:10.1007/s42979-025-04313-6 · Cited by 5
“Feature-based detectors identify unique features within an object, such as edges and corners, Methods like SIFT (Scale-Invariant Feature Transform) [12] and SURF (Speeded-Up Robust Features) [13] are employed for identifying informative regions, extracting features, and performing classification.”
Point set registration-based image stitching in unmanned aerial vehicle transmission line inspection
Wang, X., Hu, X., Xiang, D., et al. (2025) · EURASIP Journal on Advances in Signal Processing · doi:10.1186/s13634-025-01244-5 · Cited by 1
“In order to improve the algorithm speed, some improved algorithms such as ASIFT [8], PCA-SIFT [9], CSIFT [10], and SURF [11] are proposed.”
Feature Matching in the Dark: Homography-Based RGB-IR Feature Transformation for Low-Light Vision
O'Donnell, K., & Kambhamettu, C. (2025) · IEEE/CVF Conference on Computer Vision and Pattern Recognition Workshops (CVPRW) · doi:10.1109/cvprw67362.2025.00158 · Cited by 2
“Early attempts to establish correspondences between these modalities relied on adaptations of classical feature detectors and descriptors like SIFT [15], SURF [1], and ORB [17].”
Qiu, L., Liu, C., Li, G., et al. (2025) · IEEE Journal of Selected Topics in Applied Earth Observations and Remote Sensing · doi:10.1109/jstars.2025.3609808 · Cited by 1
“Basic methods rely on algorithms, such as SIFT [1], SURF [21], ORB [22], BRISK [23], and AKAZE [24], for identifying key points to align images.”
Constrained Extended Target Tracking With Random Matrix and Approximate Projection
Pei, S., Wang, Z., Li, R., et al. (2025) · IEEE Sensors Journal · doi:10.1109/jsen.2025.3609528 · Cited by 1
“There are various methods to extract feature information of targets to obtain measurements, such as Harris corner detection [46], speeded up robust features (SURFs) [47], and scale-invariant feature transform (SIFT) [48].”
Tan, Y., Yang, X., Liu, R., et al. (2025) · International Conference on Smart Grids and Energy Systems (SGES) · doi:10.1109/sges66701.2025.11155936 · Cited by 1
“Traditional image processing methods, such as edge detection, morphological operations, or feature point matching techniques (e.g., SIFT[2], SURF[3]), although effective under certain conditions, struggle with complex backgrounds, severe lighting changes, and significant geometric deformations, making their robustness…”
Collective Migration-Inspired Large-Deformation Compensation for Nonrigid Image Registration
Liu, D., Ai, D., Song, H., Fan, J., Fu, T., Xiao, D., Wang, Y., Wang, Y., & Yang, J. (2025) · International Journal of Computer Vision · doi:10.1007/s11263-025-02538-9 · Cited by 1
“Many renowned hand-crafted feature detectors and descriptors have demonstrated strong performance, such as SIFT [65], SURF [8], ORB [86], FAST [85], and BRISK [58].”
Toward Free-Form Local Feature Matching
Lu, X., Ду, С., Yan, Y., Lu, X., & Ikenaga, T. (2025) · IEEE Transactions on Pattern Analysis and Machine Intelligence · doi:10.1109/tpami.2025.3614652 · Cited by 1
Zhu, C., Mu, Y., Sun, Y., Gong, H., Guo, Y., Fan, J., Li, S., Li, Z., & Hu, T. (2025) · Agriculture · doi:10.3390/agriculture15192018 · Cited by 2
An efficient human pose keypoint detection algorithm base on YOLOv8
Yang, S., Liu, H., Du, Y., Sun, T., Wang, C., Lin, F., & Liu, L. (2025) · The Journal of Supercomputing · doi:10.1007/s11227-025-07953-w · Cited by 1
“Early representative works such as SIFT [8] and SURF [9] constructed scale-space representations and employed Hessian matrix-based keypoint detection, laying the foundation for local feature extraction.”
MDF-YOLO: A Hölder-Based Regularity-Guided Multi-Domain Fusion Detection Model for Indoor Objects
Luan, F., Yang, J., & Zhang, H. (2025) · Fractal and Fractional · doi:10.3390/fractalfract9100673 · Cited by 2
“Bay et al.’s SURF approximated Hessian responses using integral images and box filters, substantially reducing descriptor computation costs and significantly improving real-time performance while maintaining robustness [32].”
Zhang, Y., Liu, R., Zhang, Z., Shi, L., Weng, L., & Hu, L. (2025) · Remote Sensing · doi:10.3390/rs17213571 · Cited by 2
PVD: A Detection Framework for Parking Vehicles on Highways From Oblique UAV Images
Bin, J., Guo, J., Liu, G., Jiang, Z., Liu, M., Yuan, C., & Liu, X. (2025) · IEEE Journal of Selected Topics in Applied Earth Observations and Remote Sensing · doi:10.1109/jstars.2025.3629396 · Cited by 2
“Compared with common SIFT [52] and SURF [53], it has great balance between computational efficiency and matching performance.”
Palmprint Classification with Deep Learning Features and SURF Keypoints
Ignat, A., Păvăloi, I., & Niţă, C. D. (2025) · Procedia Computer Science · doi:10.1016/j.procs.2025.10.062 · Cited by 1
Fei, X., Tian, T., Choset, H., et al. (2025) · IEEE/RSJ International Conference on Intelligent Robots and Systems (IROS) · doi:10.1109/iros60139.2025.11246796 · Cited by 1
“While this approach enriches image representation by incorporating visual word pairs, BoWP relies on computationally expensive SURF features [6] and directly uses raw features as visual words, making it significantly slower compared to efficient BoW methods like DBoW2 [14].”
Visual Loop Closure Detection Through Deep Graph Consensus
Büchner, M., Dahiya, L., Dorer, S., Ramtekkar, V., Nishimiya, K., Cattaneo, D., & Valada, A. (2025) · IEEE/RSJ International Conference on Intelligent Robots and Systems (IROS) · doi:10.1109/iros60139.2025.11247208 · Cited by 1
“Visual Correspondence : Traditional methods for visual correspondence rely on handcrafted feature descriptors such as ORB [19], SIFT [20], and SURF [21].”
Zhang, Z., Song, H., Fan, J., Shao, L., Fu, T., Ai, D., Xiao, D., Wang, Y., Lin, Y., & Yang, J. (2025) · IEEE Transactions on Medical Imaging · doi:10.1109/tmi.2025.3639759 · Cited by 1
“Conventional pose estimation methods in SfM typically rely on feature extraction and matching (e.g., SIFT [26], SURF [27], or ORB [28]), which take advantage of epipolar geometry to estimate the essential matrix and decompose it to obtain the relative rotation and translation between interframes.”
Hardware-Aware Feature Extraction Quantisation for Real-Time Visual Odometry on FPGA Platforms
Wąsala, M., Smolarczyk, M., Daniłowicz, M., & Kryjak, T. (2025) · Euromicro Conference on Digital System Design (DSD) · doi:10.1109/dsd67783.2025.00080 · Cited by 3
“For VO systems that require high real-time performance, it is important to efficiently extract robust features and descriptors for matching.”
Local feature-based image matching: a comprehensive review and robustness evaluation
Bian, H., Chen, Q., Zhang, H., Qin, L., Cui, H., & Wang, X. (2025) · The Visual Computer · doi:10.1007/s00371-025-04278-w · Cited by 4
“Alcantarilla et al. [82] extended the SURF method [81] to serve as the descriptor component of the KAZE method, which constructs descriptors by calculating gradient orientations in a nonlinear scale space.”
Map Generation From Overlapping Microscopy Images Using Stitching Methods
Buță, G.-S., Ceaușescu, C.-M., & Alexe, B. (2025) · IEEE International Conference on Tools with Artificial Intelligence (ICTAI) · doi:10.1109/ictai66417.2025.00078 · Cited by 1
A Map-Based Localization System for Ingenuity Using Deep Image Matching
Georgakis, G., Pisanti, D., Williams, N., Mauceri, C., Kubiak, G., Ansar, A. I., & Brockers, R. (2025) · IEEE transactions on field robotics. · doi:10.1109/tfr.2025.3612364 · Cited by 1
“Popular hand-crafted features such as SIFT [10], ORB [29], and SURF [30] have been investigated for MbL in UAV applications.”
Stereo Vision Vibration Measurement With Occlusion-Resilient Modal Parameter Correction
Min, Q., Shi, Z., Mei, H., Zhang, Z., Zhou, Q., & Li, J. (2025) · IEEE Transactions on Instrumentation and Measurement · doi:10.1109/tim.2025.3613913 · Cited by 1
“Traditionally, several e ff ective algorithms have been developed for this purpose, including Harris [28], FAST [29], BRISK [30], FREAK [31], SIFT [32], and SURF [33].”
Jingsi, 黄. H., Siyu, 李. L., Qingwu, 段. D., & Zheng, 李. L. (2025) · Laser & Optoelectronics Progress · doi:10.3788/lop250782 · Cited by 1
Multi-feature image stitching method with edge curve matching
Yang, Z., Liu, L., Gao, T., et al. (2025) · Journal of King Saud University - Computer and Information Sciences · doi:10.1007/s44443-025-00361-3 · Cited by 1
“Single-feature-based image stitching methods typically rely on feature descriptors (Lowe 2004; Bay et al. 2006; Rublee et al. 2011) to extract point features from images and construct warping models based on the correspondences between point features.”
Yamashita, S., Kinoshita, Y., & Kiya, H. (2025) · Asia-Pacific Signal and Information Processing Association Annual Summit and Conference (APSIPA ASC) · doi:10.1109/apsipaasc65261.2025.11249232 · Cited by 1
“Homography-based approaches estimate a homography matrix between two images by leveraging feature descriptors such as SIFT [6], SURF [7], and A-KAZE [8], often in conjunction with the RANSAC algorithm [9].”
Afridi, R., Halim, M. S., Ahmed, M., Yucel, H., Fazal, Z. Z., Janjua, K., Campigotto, M., Shah, S. M., Nguyen, Q. D., & Sepah, Y. J. (2025) · Translational Vision Science & Technology · doi:10.1167/tvst.14.12.20 · Cited by 1
BFS-YOLO: a lightweight vehicle detection algorithm with a cross-layer feature fusion structure
Yang, X., Qiang, S., He, L., Gao, J., Zhu, G., Gui, C., & Ren, N. (2025) · Journal of Real-Time Image Processing · doi:10.1007/s11554-025-01835-3 · Cited by 2
DTD: Density Triangle Descriptor for 3D LiDAR Loop Closure Detection
Tang, K., Wang, Q., Yan, C., et al. (2025) · Sensors · doi:10.3390/s26010201 · Cited by 1
“These approaches typically rely on retrieving 2D image features, where classical descriptors such as SIFT [15], SURF [16], and ORB [17] offer invariance to rotation and translation.”
Autonomous Localization and Navigation for Drones Over Rivers
Wu, P.-C., Chen, Y.-Y., Lin, C.-L., et al. (2026) · IEEE Access · doi:10.1109/access.2025.3650454 · Cited by 1
“Com-VOLUME mon feature descriptors such as SIFT [8], SURF [22] or ORB [23] enable efficient and robust real-time applications.”
Li, J., Wu, J. B., Liu, B., Gong, Z., & Zhang, G. (2025) · SAE technical papers on CD-ROM/SAE technical paper series · doi:10.4271/2025-01-7306 · Cited by 1
Chen, J., & Cao, Y. (2026) · Engineering Construction & Architectural Management · doi:10.1108/ecam-07-2025-1095 · Cited by 1
Lee, K., Lee, S., & Lee, K. M. (2026) · IEEE Transactions on Image Processing · doi:10.1109/tip.2026.3653198 · Cited by 1
“Point correspondence search meth-ods [5, 6, 7, 8, 9, 10] or global transformation estimation methods [11, 12, 13, 14] struggle when distinctive features are sparse.”
Deep learning for object detection: state of the art, challenges, and future directions
Gupta, N., & Rohil, M. K. (2026) · International Journal of Data Science and Analysis · doi:10.1007/s41060-026-01030-y · Cited by 3
“Feature detectors like scale invariant feature transform (SIFT) [21], speeded-up robust features (SURF) [22], Haar-like features [23], etc., are often used to perform this desired task; however, diversity of imaging conditions and appearances makes it difficult to correctly determine all kinds of…”
Wang, Z., Zuo, X., & Dong, W. (2026) · IEEE Transactions on Robotics · doi:10.1109/tro.2026.3658293 · Cited by 3
“RANSAC [76] (ORB_NR), and SURF [38] + NN matching + RANSAC (SURF_NR).”
MambaSC: A Feature Matching Method Using Mamba2 with Self and Cross-Attention for Multimodal Images
Teng, R., Liao, Y., Wang, W., Duan, Q., Liu, J., & Jin, F. (2026) · Journal of Artificial Intelligence and Soft Computing Research · doi:10.2478/jaiscr-2026-0009 · Cited by 1
“Classical approaches such as SIFT [1], SURF [2], and ORB [3] detect keypoints invariant to scale, rotation, or illumination, enabling robust matching across view-points.”
Real-Time FPGA-Based Hardware-Algorithm Co-Design for Monocular Visual Odometry on Edge Devices
Huang, L.-Y., Hsieh, Y.-K., & Chien, S. (2026) · IEEE Transactions on Circuits and Systems for Video Technology · doi:10.1109/tcsvt.2026.3664685 · Cited by 1
“The first stage involves the detection of keypoints and establishing feature correspondences using handcrafted features such as SIFT [11], ORB [12], and SURF [13].”
Zhou, L., Peng, T., Zhiqiang, H. A. N., Li, L., Zhu, Q. Q., & Ye, Y. (2026) · IEEE Transactions on Geoscience and Remote Sensing · doi:10.1109/tgrs.2026.3664764 · Cited by 4
“Algorithms like scale invariance feature transform (SIFT) [29] and speeded-up robust features (SURFs) [30] can extract descriptors that are invariant to scale and rotation, thereby overcoming geometric changes to some extent.”
Autonomous robotic exploration of unknown soft objects
Hu, J., Jones, D., Galindo, G. L., Huang, S., & Valdastri, P. (2026) · The International Journal of Robotics Research · doi:10.1177/02783649251415415 · Cited by 6
“Both approaches fundamentally depend on establishing correspondences across images, often achieved using feature detection techniques such as SURF (Bay et al., 2006) or motion tracking methods like scene fl ow (Chen et al., 2024).”
A Robotic Grasping Framework for Railway Cargo Based on an Improved YOLOv8 Model
Xiong, S., Wu, H., & Lin, Y. (2025) · doi:10.1109/icrsa66467.2025.11383796 · Cited by 1
“Subsequently, feature point detection methods such as SIFT (Scale-Invariant Feature Transform) [13], SURF (Speeded-Up Robust Features) [14], and ORB (Oriented FAST and Rotated BRIEF) [15] are employed to extract keypoints and match them with template images, establishing 2D-2D correspondences.”
Zhang, C., Zhang, Y., Li, L., et al. (2026) · Sensors · doi:10.3390/s26041324 · Cited by 1
Arakawa, S., Suzuki, K., & Matsuzawa, T. (2026) · Sensors · doi:10.3390/s26041322 · Cited by 1
Similarity-Aware Selective State-Space Modeling for Semantic Correspondence
Kim, S., & Cho, M. (2025) · IEEE International Conference on Computer Vision Workshops · doi:10.1109/iccvw69036.2025.00641 · Cited by 1
“Traditional methods [2, 4, 13, 32] typically use handcrafted descriptors [1, 7, 35] which are simple yet effective.”
Nakao, T., Hara, Y., & KURODA, Y. (2026) · IEEE/SICE International Symposium on System Integration (SII) · doi:10.1109/sii64115.2026.11404603 · Cited by 1
“Classical VPR methods extended image retrieval tasks by representing local descriptors, such as SIFT [10] and SURF [11], as feature vectors using Bag of Visual Words (BoVW) [12] or VLAD [13].”
Planar-Guided Gaussian Splatting with Texture-Complexity-Based Initialization
Zheng, A., & Yu, Z. (2026) · Electronics · doi:10.3390/electronics15051137 · Cited by 1
Towards Pseudo-Labeling with Dynamic Thresholds for Cross-View Image Geolocalization
Yuan, Y., Guo, J., Zhu, R., Li, N., Li, Z., & Luo, W. (2026) · Remote Sensing · doi:10.3390/rs18060944 · Cited by 1
Textureless Surface Feature Point Detection via Micro-Geometry Reconstruction
Liang, Y., Wang, Y., Yan, T., Yang, J., Li, W., Huang, L., Ning, X., & Ataxonovich, K. T. (2026) · IEEE Transactions on Pattern Analysis and Machine Intelligence · doi:10.1109/tpami.2026.3681931 · Cited by 1
“(cid:2) Classical gradient-based methods: Shi-Tomasi [15], SIFT [5], SURF [16], FAST [6], ORB [7], BRISK [17], KAZE [8], AKAZE [18].”
MIFTr: Efficient Universal Image Matching With Modality-Invariant Feature Transformer
Gao, C., Gao, Y., Weng, D., Li, J., Li, W., Tao, R., & Xia, X. (2026) · IEEE Transactions on Geoscience and Remote Sensing · doi:10.1109/tgrs.2026.3683096 · Cited by 3
“Starting from the earliest SIFT [17], [18], [19], the traditional approaches [20], [21], [22], [23], [24], [25], [26], [27], [28] have undergone decades of development and become highly established.”
Robust Scene Coordinate Regression via Geometrically-Consistent Global Descriptors
Nguyen, S. T., Fontan, A., Milford, M. J., & Fischer, T. (2026) · IEEE Winter Conference on Applications of Computer Vision · doi:10.1109/wacv61042.2026.00792 · Cited by 1
“Classical methods [4, 37] detect invariant keypoints that can be reliably matched across different views.”
GDoFS: Gaussian DoF Separation for Plausible 3D Geometry in Sparse-View 3DGS
Kim, Y., Choi, J., & Yoon, S. (2026) · IEEE Winter Conference on Applications of Computer Vision · doi:10.1109/wacv61042.2026.00642 · Cited by 1
A lifelong mapping framework with multi-view projection in a changing environment
Xian, K., Liu, D., Zhang, G., Neri, F., & Chen, S. (2026) · Integrated Computer-Aided Engineering · doi:10.1177/10692509261442329 · Cited by 1
“Classical feature-based methods, such as Scale-Invariant Feature Transform (SIFT), 40 Speeded Up Robust Features (SURF), 41,42 and Oriented FAST and Rotated BRIEF (ORB), 43 have demonstrated robust performance in image matching and are widely adopted in visual SLAM systems.”
RaCo: Ranking and Covariance for Practical Learned Keypoints
Shenoi, A., Lindenberger, P., Sarlin, P.-E., & Pollefeys, M. (2026) · International Conference on 3D Vision (3DV) · doi:10.1109/3dv69130.2026.00033 · Cited by 4
“Decoupling keypoints and descriptors was predominant in the era of handcrafted local-features [6, 40, 48], which followed a detect-then-describe paradigm.”
Are We Ready to Enable Satellite Autonomy Through on Board Image Registration?
Salvo, C. D., Sozzo, E. D., Sorrentino, G., D’Arnese, E., Panicucci, P., & Conficconi, D. (2026) · Proceedings of the ACM on Measurement and Analysis of Computing Systems · doi:10.1145/3805628 · Cited by 2
“To establish a baseline for computational overhead versus robustness, we additionally benchmark other methods such as SIFT [36], SURF [6], and ORB [47].”
Evaluating Local Feature Detectors in Salient Region Detection
Kinnunen, T., Laine‐Hernandez, M., & Oittinen, P. T. (2013) · Lecture notes in computer science · doi:10.1007/978-3-642-38886-6_9 · Cited by 1
Combining Image Similarity Metrics for Semantic Image Annotation
Jansen, B., Toan, T. D., & Temmermans, F. (2012) · Lecture notes in computer science · doi:10.1007/978-3-642-33618-8_40 · Cited by 1
Freight Status Classification in Real-World Images Using SIFT and KNN Model
Wang, D., Yu, D., Han, J., et al. (2013) · Lecture notes in electrical engineering · doi:10.1007/978-3-319-00536-2_17 · Cited by 1
Evolving Local Descriptor Operators through Genetic Programming
Pérez, C. B., & Olague, G. (2009) · Lecture notes in computer science · doi:10.1007/978-3-642-01129-0_46 · Cited by 1
A SURF Feature Based Building Recognition System for Distinctive Architectures
Chen, S.-W., Chung, Y.-H., Chien, H.-F., et al. (2013) · Lecture notes in electrical engineering · doi:10.1007/978-94-007-6996-0_12 · Cited by 1
Moving Object Detection Using Monocular Vision
Wang, Y.-T., Chen, K.-W., & Chiou, M.-J. (2012) · Advances in intelligent systems and computing · doi:10.1007/978-3-642-33926-4_17 · Cited by 1
A Localization Framework under Non-rigid Deformation for Robotic Surgery
Xiang, X. (2011) · Lecture notes in computer science · doi:10.1007/978-3-642-24028-7_2 · Cited by 1
Accuracy of Trajectories Estimation in a Driver-Assistance Context
Khan, W., & Klette, R. (2014) · Lecture notes in computer science · doi:10.1007/978-3-642-53926-8_5 · Cited by 1
“The radius of this neighbourhood is derived from the scale at which the FP was detected [3].”
BackgroundMutual Information Refinement for Flash-no-Flash Image Alignment
Varjo, S. J. O., Hannuksela, J., Sílven, O., et al. (2011) · Lecture notes in computer science · doi:10.1007/978-3-642-23687-7_37 · Cited by 1
“Other well known methods for interest points and descriptors are the Smallest Univalue Segment Assimilating Nuclei (SUSAN) [23], the Scale Invariant Feature Transform (SIFT) [15] and the Speeded Up Robust Features (SURF) [3].”
MethodAccurate Face and Human Detection Using Hybrid Local Transform Features
Kim, D., & Jun, B. (2015) · KAIST research series · doi:10.1007/978-94-017-9987-4_8 · Cited by 1
Building Detection and 3D Reconstruction from Two-View of Monocular Camera
Le, M.-H., & Jo, K.-H. (2011) · Lecture notes in computer science · doi:10.1007/978-3-642-23935-9_42 · Cited by 1
“There are many kind of features are considered in recent research in feature extraction and matching include Harris [31], SIFT, PCA-SIFT, SURF [32], [33], etc.”
MethodEvaluation of Interest Point Detectors for Non-planar, Transparent Scenes
Papalazarou, C., Rongen, P. M. J., & de With, P. H. N. (2009) · Lecture notes in computer science · doi:10.1007/978-3-642-04697-1_1 · Cited by 1
Stereo Vision and Its Application to Robotic Manipulation
Takamatsu, J. (2011) · InTech eBooks · doi:10.5772/23399 · Cited by 1
Log-Spiral Keypoint: A Robust Approach toward Image Patch Matching
Paek, K., Yao, M., Liu, Z., et al. (2015) · Computational Intelligence and Neuroscience · doi:10.1155/2015/457495 · Cited by 1
Monocular optical flow navigation using sparse SURF flow with multi-layer bucketing screener
Qian, C., Wang, Y., & Guo, L. (2015) · Chinese Control Conference (CCC) · doi:10.1109/chicc.2015.7260225 · Cited by 1
“However, Herbert noted in [10] that SURF run several times faster than SIFT and still remains the similar performance.”
BackgroundMethodA Novel Binary Feature from Intensity Difference Quantization between Random Sample of Points
Zhuang, D., Zhang, D., Li, J., et al. (2013) · Lecture notes in computer science · doi:10.1007/978-3-642-35728-2_22 · Cited by 1
“SURF (Speeded-up Robust Feature) [2] and GLOH (Gradient Location Orientation Histogram) [8]) are proposed, but these approaches still can not finally resolve the conflict between efficiency and effectiveness.”
BackgroundHybrid Document Matching Method for Page Identification of Digilog Books
Park, J.-H., & Woo, W. (2011) · Lecture notes in computer science · doi:10.1007/978-3-642-23456-9_1 · Cited by 1
Explore Objects and Categories in Unexplored Environments Based on Multimodal Data
Zhang, J. (2012) · Cited by 1
Dayoub, F., Cielniak, G., & Duckett, T. (2014) · Springer tracts in advanced robotics · doi:10.1007/978-3-319-07488-7_26 · Cited by 1
Using a Markov Random Field for Image Re-ranking Based on Visual and Textual Features
Chavéz-García, R. O., Montes, M., & Sucar, L. E. (2011) · Computación y Sistemas · doi:10.13053/cys-14-4-1284 · Cited by 1
RANSAC-Based Stereo Image Registration with Geometrically Constrained Hypothesis Generation
Cupec, R., Nyarko, E. K., Kitanov, A., & Petrovic, I. K. (2009) · Automatika · Cited by 1
An approach for combining multiple descriptors for image classification
Tran, D. T., Jansen, B., Deklerck, R., & Debeir, O. (2015) · Proceedings of SPIE, the International Society for Optical Engineering/Proceedings of SPIE · doi:10.1117/12.2181017 · Cited by 1
“In our implementation, two out of four image descriptors (GB and SURF) outperform previous results (including NN-based method or other approaches).”
InfluentialReconocimiento de objetos en video utilizando SIFT paralelo
Albanesi, M. B., Funes, N., Chichizola, F., et al. (2010) · El Servicio de Difusión de la Creación Intelectual (National University of La Plata) · Cited by 1
A new ranking approach and a revisited ratio test for improving content-based image retrieval
Mauro, M., & Leonardi, R. (2012) · Proceedings - International Conference on Image Processing · doi:10.1109/icip.2012.6467383 · Cited by 1
Comparison of Classifiers for Retinal Pathology Images using SURF and Bag-of- Words Model
Mukti, F. A., & Hashim, N. (2014) · International Conference on Artificial Intelligence · Cited by 1
On user-defined region matching for augmented reality
Zhou, J., & Payandeh, S. (2015) · IEEE Canadian Conference on Electrical and Computer Engineering (CCECE) · doi:10.1109/ccece.2015.7129487 · Cited by 1
Hand Posture Recognition with Multiview Descriptors
Collumeau, J.-F., Laurent, H., Emile, B., et al. (2012) · Lecture notes in computer science · doi:10.1007/978-3-642-33140-4_40 · Cited by 1
New Hierarchical Approaches in Real-Time Robust Image Feature Detection and Matching
Langer, M., & Kuhnert, K. ‐. (2008) · InTech eBooks · doi:10.5772/6163 · Cited by 1
Video shots annotation using random forest
Cai, C., & Li, Z. (2015) · Asian Control Conference (ASCC) · doi:10.1109/ascc.2015.7244736 · Cited by 1
“The features we extracted from multi-keyframe are Gist [5], HOG [6], SURF [7], as well as the combinations of these features, and each feature should be normalized by (1).”
MethodInfluentialProfile face detection using flipping scheme with genetic algorithm
You, M., & Akashi, T. (2015) · Asian Control Conference (ASCC) · doi:10.1109/ascc.2015.7244748 · Cited by 1
“SURF (Speeded Up Robust Features [11]) is a local scaleand rotation-invariant interest point detector and descriptor defined in the gradient space.”
BackgroundEngelbrecht, K.-P., Schmiedeke, S., Quade, M., & Möller, S. (2015) · International Conference on Intelligent Environments (IE) · doi:10.1109/ie.2015.9 · Cited by 1
“Thus, we apply a combined method using background modeling [16] and SURF [1] for generating keypoints with a potential high chance of right correspondences.”
Mobile Vision Based Augmented Reality Navigation System
Wojciechowski, A. (2012) · TU repository (Lodz University of Technology) · Cited by 1
Top down scene context based visual attention model for natural images
Andrushia, A. D., & Thangarajan, R. (2015) · International Conference on Electronics and Communication Systems (ICECS) · doi:10.1109/ecs.2015.7124971 · Cited by 1
Zhao, B., Hu, T., Niu, Y., Tang, D., Ma, Z., Kong, W., & Shen, L. (2015) · IEEE International Conference on Information and Automation (ICIA) · doi:10.1109/icinfa.2015.7279258 · Cited by 1
A Fast Machine Vision for Automated Packaging
Liu, Y., & Vo, T. V. (2012) · Lecture notes in electrical engineering · doi:10.1007/978-1-4471-2467-2_85 · Cited by 1
Scalable Video Genre Classification and Event Detection
Muneesawang, P., Zhang, N., & Guan, L. (2014) · Springer eBooks · doi:10.1007/978-3-319-11782-9_9 · Cited by 2
Rapid target recognition and tracking under large scale variation using Semi-Naive Bayesian
Sun, K., Wang, B., & Hao, Z. (2010) · Cited by 1
Hierarchical Image Representation Using Deep Network
Ergül, E., Ertürk, S., & Arıca, N. (2015) · Lecture notes in computer science · doi:10.1007/978-3-319-23234-8_7 · Cited by 1
“Unlike hand-engineered feature extraction methods like Scale Invariant Feature Transform (SIFT) [3], Speeded Up Robust Features (SURF) [4], Pyramid Histogram of Oriented Gradients (PHOG) [5] or GIST [6]; DLA can extract useful features without explicitly using input data statistics.”
MethodSpatio-temporal Features for Efficient Video Copy Detection
Hu, R., Li, B., Hu, W., & Yang, J. (2013) · Lecture notes in computer science · doi:10.1007/978-3-642-42057-3_16 · Cited by 1
“Cameron first extracts SURF features [5] for the frames and then creates temporal signature by sorting the SURF feature counts in each region along the time-line.”
Matching of Feature Points Based on TSSC Method from MR Images of Nonrigid Deformed Tissues
Zhang, X., Zhang, P., & Hirai, S. (2014) · Bio-Medical Materials and Engineering · doi:10.3233/bme-130924 · Cited by 1
“The Speeded-UP Robust Feature (SURF) method [20] was shown to be more reproducible, distinctive, and robust than SIFT, and to experimentally outperform other methods, such as GLOH and PCA-SIFT [19], [20-21].”
lMAR: Highly Parallel Architecture for Markerless Augmented Reality in Aircraft Maintenance
Caponio, A., Hincapié, M., & Mendívil, E. G. (2011) · Lecture notes in computer science · doi:10.1007/978-3-642-22021-0_3 · Cited by 1
Towards Intelligent Systems with Incremental Learning Ability
Reiff, T., & Sinčák, P. (2009) · Studies in computational intelligence · doi:10.1007/978-3-642-03737-5_45 · Cited by 1
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.