Treffer: Method for estimating real-scale 3D human body shape from an image based on 3D camera calibration and computer graphics-based reverse projection photogrammetry.
Title:
Method for estimating real-scale 3D human body shape from an image based on 3D camera calibration and computer graphics-based reverse projection photogrammetry.
Authors:
Imoto D; Artificial Intelligence Section, Second Department of Forensic Science, National Research Institute of Police Science, Chiba, Japan., Honma M; Artificial Intelligence Section, Second Department of Forensic Science, National Research Institute of Police Science, Chiba, Japan., Asano M; Artificial Intelligence Section, Second Department of Forensic Science, National Research Institute of Police Science, Chiba, Japan., Sakurai W; Artificial Intelligence Section, Second Department of Forensic Science, National Research Institute of Police Science, Chiba, Japan., Kurosawa K; Second Department of Forensic Science, National Research Institute of Police Science, Chiba, Japan.
Source:
Journal of forensic sciences [J Forensic Sci] 2025 Nov; Vol. 70 (6), pp. 2276-2296. Date of Electronic Publication: 2025 Jul 06.
Publication Type:
Journal Article
Language:
English
Journal Info:
Publisher: Blackwell Pub Country of Publication: United States NLM ID: 0375370 Publication Model: Print-Electronic Cited Medium: Internet ISSN: 1556-4029 (Electronic) Linking ISSN: 00221198 NLM ISO Abbreviation: J Forensic Sci Subsets: MEDLINE
Imprint Name(s):
Publication: 2006- : Malden, MA : Blackwell Pub.
Original Publication: [Chicago, Ill.] : Callaghan and Co., 1956-
Original Publication: [Chicago, Ill.] : Callaghan and Co., 1956-
MeSH Terms:
References:
Imoto D, Kurosawa K, Tsuchiya K, Kuroki K, Hirabayashi M, Akiba N, et al. A novel human identification method by gait using dynamics of feature points and local shape features. Proceedings of the 2018 7th European Workshop on Visual Information Processing (EUVIP); 2018 Nov 26–28; Tampere, Finland. Piscataway, NJ: IEEE; 2018. p. 1–6.
Imoto D, Kurosawa K, Honma M, Yokota R, Hirabayashi M, Hawai Y. Model‐based interpolation for continuous human silhouette images by height‐constraint assumption. Proceedings of the 2020 4th international conference on vision, image and signal processing (ICVISP); 2020 Dec 9–11; Bangkok, Thailand. New York, NY: ACM; 2020. p. 1–11.
Imoto D, Hirabayashi M, Honma M, Kurosawa K. Enhancing the robustness of forensic gait analysis against near‐distance viewing direction differences. Multimed Tools Appl. 2022;81:26199–26221. https://doi.org/10.1007/s11042‐022‐12751‐0.
Imoto D, Hirabayashi M, Honma M, Kurosawa K. Pre‐set estimation‐based in‐silico silhouette‐based methodology for improving the robustness to viewing direction difference for assisting forensic gait analysis. J Forensic Sci. 2023;68(2):470–487. https://doi.org/10.1111/1556‐4029.15214.
Imoto D, Asano M, Sakurai W, Honma M, Kurosawa K. Running gait biometrics at a distance: A novel silhouette/3d running gait dataset in forensic scenarios. Proceedings of the 23rd International Conference of Biometrics Special Interest Group (IEEE BIOSIG); 2024 Sep 26–27; Darmstadt, Germany. Piscataway, NJ: IEEE; 2024. p. 1–7.
Hartley RI. In defense of the eight‐point algorithm. IEEE Trans on Pattern Anal and Mach Intell (T‐PAMI). 1997;19(6):580–593. https://doi.org/10.1109/34.601246.
Hartley RI, Zisserman A. The direct linear transform (DLT) algorithm. Multiple view geometry in computer vision. Cambridge: Cambridge University Press; 2004. p. 88–93.
Lepetit V, Moreno‐Noguer F, Fua P. EPnP: an accurate O(n) solution to the PnP problem. Int J Comput vis. 2009;81:155–166. https://doi.org/10.1007/s11263‐008‐0152‐6.
Chen H, Tian W, Wang P, Wang F, Xiong L, Li H. EPro‐PnP: generalized end‐to‐end probabilistic perspective‐n‐points for monocular object pose estimation. IEEE Trans Pattern Anal Mach Intell. 2020;20:997.
Agarwal S, Snavely N, Simon I, Seitz SM, Szeliski R. Building rome in a day. Proceedings of the 2009 IEEE 12th international conference on computer vision (ICCV); 2009 Sep 27‐Oct 4; Kyoto, Japan. Piscataway, NJ: IEEE; 2009. p. 72–79.
Berghi D, Cieciura C, Einabadi F, Glancy M, Camilleri OC, Foster PA, et al. ForecasterFlexOBM: a multi‐view audio‐visual dataset for flexible object‐based media production. Proceedings of the 2024 IEEE International Conference on Multimedia and Expo; 2024 Jul 15–19; Niagara Falls, Canada. Piscataway, NJ: IEEE; 2024. p. 1–6.
Park G, Lee JH, Yoon H. Semantic structure from motion for railroad bridges using deep learning. Appl Sci. 2021;11(10):4332. https://doi.org/10.3390/app11104332.
Ospina‐Bohórquez A, Del Pozo S, Courtenay LA, González‐Aguilera D. Handheld stereo photogrammetry applied to crime scene analysis. Measurement. 2023;216:112861. https://doi.org/10.1016/j.measurement.2023.112861.
Raneri D. Enhancing forensic investigation through the use of modern three‐dimensional (3D) imaging technologies for crime scene reconstruction. Aust J Forensic Sci. 2018;50(6):697–707. https://doi.org/10.1080/00450618.2018.1424245.
Galanakis G, Allertseder S, Zabulis X, Evdaimon T, Fikenscher SE, Tsikrika T, et al. A study of 3D digitisation modalities for crime scene investigation. Forensic Sci. 2021;1:56–85. https://doi.org/10.3390/forensicsci1020008.
Desmoulina GT, Kalkata M, Milner TE. Forensic application of inverse and reverse projection photogrammetry to determine subject location and orientation when both camera and subject move relative to the scene. Forensic Sci Int. 2022;331:111145. https://doi.org/10.1016/j.forsciint.2021.111145.
Imoto D, Honma M, Akiba N, Hirabayashi M, Onozuka S, Akita K, et al. Development for geometric image analysis programs for forensic engineering examination. Jpn J Forensic Sci Technol. 2023;28(1):15–42. (in Japanese). https://doi.org/10.3408/jafst.833.
Kikkawa S, Okura F, Muramatsu D, Yagi Y, Saito H. Accuracy evaluation and prediction of single‐image camera calibration. IEEE Access. 2023;11:19312–19323. https://doi.org/10.1109/ACCESS.2023.3244212.
Imoto D, Kurosawa K, Tsuchiya K, Kuroki K, Akiba N, Kakuda H. Error estimation of indirect length measurement from an image based on the law of cross‐ratio preservation. Rep Natl Res Inst Police Sci. 2018;67(2):35–41. (in Japanese).
Hoogeboom B, Alberink I, Goos M. Body height measurements in images. J Forensic Sci. 2009;54(6):1365–1375. https://doi.org/10.1111/j.1556‐4029.2009.01179.x.
Edelman G, Alberink I, Hoogeboom B. Comparison of the performance of two methods for height estimation. J Forensic Sci. 2010;55(2):358–365. https://doi.org/10.1111/j.1556‐4029.2009.01296.x.
Johnson M, Liscio E. Suspect height estimation using the faro focus3d laser scanner. J Forensic Sci. 2015;60(6):1582–1588. https://doi.org/10.1111/1556‐4029.12829.
Liscio E, Guryn H, Le Q, Olver A. A comparison of reverse projection and photomodeler for suspect height analysis. Forensic Sci Int. 2021;320:110690. https://doi.org/10.1016/j.forsciint.2021.110690.
Tosti F, Nardinocchi C, Wahbeh W, Ciampini C, Marsella M, Lopes P, et al. Human height estimation from highly distorted surveillance image. J Forensic Sci. 2021;67(1):322–344. https://doi.org/10.1111/1556‐4029.14888.
Ciampini C, Petrillo A, Zomparelli F, Groutas S. An innovative method for human height estimation combining video images and 3D laser scanning. J Forensic Sci. 2023;69(1):301–315. https://doi.org/10.1111/1556‐4029.15378.
Liscio E, Lim J. Inter‐observer variation of head and foot point selection for subject height determination. J Forensic Sci. 2024;69(4):1268–1288. https://doi.org/10.1111/1556‐4029.15529.
Loper M, Mahmood N, Romero J, Pons‐Moll G, Black MJ. SMPL: A skinned multi‐person linear model. ACM Trans Graph. 2015;34(6):248. https://doi.org/10.1145/2816795.281801.
Pavlakos G, Choutas V, Ghobani N, Bolkart T, Osman AAA, Tzionas D, et al. Expressive body capture: 3D hands, face, and body from a single image. Proceedings of the 2019 IEEE/CVF International Conference on Computer Vision and Pattern Recognition (CVPR); 2019 June 16–20; Long Beach, CA. IEEE: Piscataway, NJ; 2024. p. 10967–10977.
Thakkar N, Farid H. On the feasibility of 3d model‐based forensic height and weight estimation. Proceedings of the IEEE Conference on Computer Vision and Pattern Recogognition Workshop (CVPRW); 2021 June 19–25; held virtually. Piscataway, NJ: IEEE; 2021. p. 953–961.
Thakkar N, Pavlakos G, Farid H. The reliability of forensic body‐shape identification. Proceedings of the IEEE Conference on Computer Vision and Pattern Recogognition Workshop (CVPRW); 2022 June 19–24; New Orleans, LA. Piscataway, NJ: IEEE; 2022. p. 44–52.
Goel S, Pavlakos G, Rajasegaran J, Kanazawa A, Marik J. Humans in 4D: Reconstructing and tracking humans in transformers. Proceedings of the 2023 IEEE/CVF International Conference on Computer Vision (ICCV); 2023 October 2–6; Paris, France. Piscataway, NJ: IEEE; 2023. p. 14737–14748.
Agisoft. Metashape: Standard edition. https://www.agisoft.com/features/standard‐edition/. Accessed 31 Jan 2024.
Blender. Blender – A 3D modelling and rendering package. http://www.blender.org. Accessed 31 Jan 2024.
Meshlab. https://www.meshlab.net/. Accessed 31 Jan 2024.
CloudCompare. CloudCompare – 3D point cloud and mesh processing software Open Source Project. https://www.danielgm.net/cc/. Accessed 31 Jan 2024.
Devernay F, Faugeras O. Straight lines have to be straight. Mach Vis Appl. 2001;13:14–24. https://doi.org/10.1007/PL00013269.
Leica Geosystems. Leica FlexLine TS03 manual total station. https://leica‐geosystems.com/en‐us/products/total‐stations/manual‐total‐stations/leica‐flexline‐ts03. Accessed 31 Jan 2024.
Rajasegaran J, Pavlakos G, Kanazawa A, Malik J. Tracking people by predicting 3D appearance, location and pose. Proceedings of the IEEE Conference on Computer Vision and Pattern Recognition (CVPR); 2022 June 19–24; New Orleans, LA. Piscataway, NJ: IEEE; 2022. p. 2730–2739.
Bertillon A, McClaughry RW. Signaletic instructions including the theory and practice of anthro‐pometrical identification. Whitefish, MT: Kessinger Publishing; 1896.
Heuschkel ML, Labudde D. Reconsideration of bertillonage in the age of digitalisation: digital anthropometric patterns as a promising method for establishing identity. Forensic Sci Int Synergy. 2024;8:1004524. https://doi.org/10.1016/j.fsisyn.2023.100452.
Imoto D, Kurosawa K, Honma M, Yokota R, Hirabayashi M, Hawai Y. Model‐based interpolation for continuous human silhouette images by height‐constraint assumption. Proceedings of the 2020 4th international conference on vision, image and signal processing (ICVISP); 2020 Dec 9–11; Bangkok, Thailand. New York, NY: ACM; 2020. p. 1–11.
Imoto D, Hirabayashi M, Honma M, Kurosawa K. Enhancing the robustness of forensic gait analysis against near‐distance viewing direction differences. Multimed Tools Appl. 2022;81:26199–26221. https://doi.org/10.1007/s11042‐022‐12751‐0.
Imoto D, Hirabayashi M, Honma M, Kurosawa K. Pre‐set estimation‐based in‐silico silhouette‐based methodology for improving the robustness to viewing direction difference for assisting forensic gait analysis. J Forensic Sci. 2023;68(2):470–487. https://doi.org/10.1111/1556‐4029.15214.
Imoto D, Asano M, Sakurai W, Honma M, Kurosawa K. Running gait biometrics at a distance: A novel silhouette/3d running gait dataset in forensic scenarios. Proceedings of the 23rd International Conference of Biometrics Special Interest Group (IEEE BIOSIG); 2024 Sep 26–27; Darmstadt, Germany. Piscataway, NJ: IEEE; 2024. p. 1–7.
Hartley RI. In defense of the eight‐point algorithm. IEEE Trans on Pattern Anal and Mach Intell (T‐PAMI). 1997;19(6):580–593. https://doi.org/10.1109/34.601246.
Hartley RI, Zisserman A. The direct linear transform (DLT) algorithm. Multiple view geometry in computer vision. Cambridge: Cambridge University Press; 2004. p. 88–93.
Lepetit V, Moreno‐Noguer F, Fua P. EPnP: an accurate O(n) solution to the PnP problem. Int J Comput vis. 2009;81:155–166. https://doi.org/10.1007/s11263‐008‐0152‐6.
Chen H, Tian W, Wang P, Wang F, Xiong L, Li H. EPro‐PnP: generalized end‐to‐end probabilistic perspective‐n‐points for monocular object pose estimation. IEEE Trans Pattern Anal Mach Intell. 2020;20:997.
Agarwal S, Snavely N, Simon I, Seitz SM, Szeliski R. Building rome in a day. Proceedings of the 2009 IEEE 12th international conference on computer vision (ICCV); 2009 Sep 27‐Oct 4; Kyoto, Japan. Piscataway, NJ: IEEE; 2009. p. 72–79.
Berghi D, Cieciura C, Einabadi F, Glancy M, Camilleri OC, Foster PA, et al. ForecasterFlexOBM: a multi‐view audio‐visual dataset for flexible object‐based media production. Proceedings of the 2024 IEEE International Conference on Multimedia and Expo; 2024 Jul 15–19; Niagara Falls, Canada. Piscataway, NJ: IEEE; 2024. p. 1–6.
Park G, Lee JH, Yoon H. Semantic structure from motion for railroad bridges using deep learning. Appl Sci. 2021;11(10):4332. https://doi.org/10.3390/app11104332.
Ospina‐Bohórquez A, Del Pozo S, Courtenay LA, González‐Aguilera D. Handheld stereo photogrammetry applied to crime scene analysis. Measurement. 2023;216:112861. https://doi.org/10.1016/j.measurement.2023.112861.
Raneri D. Enhancing forensic investigation through the use of modern three‐dimensional (3D) imaging technologies for crime scene reconstruction. Aust J Forensic Sci. 2018;50(6):697–707. https://doi.org/10.1080/00450618.2018.1424245.
Galanakis G, Allertseder S, Zabulis X, Evdaimon T, Fikenscher SE, Tsikrika T, et al. A study of 3D digitisation modalities for crime scene investigation. Forensic Sci. 2021;1:56–85. https://doi.org/10.3390/forensicsci1020008.
Desmoulina GT, Kalkata M, Milner TE. Forensic application of inverse and reverse projection photogrammetry to determine subject location and orientation when both camera and subject move relative to the scene. Forensic Sci Int. 2022;331:111145. https://doi.org/10.1016/j.forsciint.2021.111145.
Imoto D, Honma M, Akiba N, Hirabayashi M, Onozuka S, Akita K, et al. Development for geometric image analysis programs for forensic engineering examination. Jpn J Forensic Sci Technol. 2023;28(1):15–42. (in Japanese). https://doi.org/10.3408/jafst.833.
Kikkawa S, Okura F, Muramatsu D, Yagi Y, Saito H. Accuracy evaluation and prediction of single‐image camera calibration. IEEE Access. 2023;11:19312–19323. https://doi.org/10.1109/ACCESS.2023.3244212.
Imoto D, Kurosawa K, Tsuchiya K, Kuroki K, Akiba N, Kakuda H. Error estimation of indirect length measurement from an image based on the law of cross‐ratio preservation. Rep Natl Res Inst Police Sci. 2018;67(2):35–41. (in Japanese).
Hoogeboom B, Alberink I, Goos M. Body height measurements in images. J Forensic Sci. 2009;54(6):1365–1375. https://doi.org/10.1111/j.1556‐4029.2009.01179.x.
Edelman G, Alberink I, Hoogeboom B. Comparison of the performance of two methods for height estimation. J Forensic Sci. 2010;55(2):358–365. https://doi.org/10.1111/j.1556‐4029.2009.01296.x.
Johnson M, Liscio E. Suspect height estimation using the faro focus3d laser scanner. J Forensic Sci. 2015;60(6):1582–1588. https://doi.org/10.1111/1556‐4029.12829.
Liscio E, Guryn H, Le Q, Olver A. A comparison of reverse projection and photomodeler for suspect height analysis. Forensic Sci Int. 2021;320:110690. https://doi.org/10.1016/j.forsciint.2021.110690.
Tosti F, Nardinocchi C, Wahbeh W, Ciampini C, Marsella M, Lopes P, et al. Human height estimation from highly distorted surveillance image. J Forensic Sci. 2021;67(1):322–344. https://doi.org/10.1111/1556‐4029.14888.
Ciampini C, Petrillo A, Zomparelli F, Groutas S. An innovative method for human height estimation combining video images and 3D laser scanning. J Forensic Sci. 2023;69(1):301–315. https://doi.org/10.1111/1556‐4029.15378.
Liscio E, Lim J. Inter‐observer variation of head and foot point selection for subject height determination. J Forensic Sci. 2024;69(4):1268–1288. https://doi.org/10.1111/1556‐4029.15529.
Loper M, Mahmood N, Romero J, Pons‐Moll G, Black MJ. SMPL: A skinned multi‐person linear model. ACM Trans Graph. 2015;34(6):248. https://doi.org/10.1145/2816795.281801.
Pavlakos G, Choutas V, Ghobani N, Bolkart T, Osman AAA, Tzionas D, et al. Expressive body capture: 3D hands, face, and body from a single image. Proceedings of the 2019 IEEE/CVF International Conference on Computer Vision and Pattern Recognition (CVPR); 2019 June 16–20; Long Beach, CA. IEEE: Piscataway, NJ; 2024. p. 10967–10977.
Thakkar N, Farid H. On the feasibility of 3d model‐based forensic height and weight estimation. Proceedings of the IEEE Conference on Computer Vision and Pattern Recogognition Workshop (CVPRW); 2021 June 19–25; held virtually. Piscataway, NJ: IEEE; 2021. p. 953–961.
Thakkar N, Pavlakos G, Farid H. The reliability of forensic body‐shape identification. Proceedings of the IEEE Conference on Computer Vision and Pattern Recogognition Workshop (CVPRW); 2022 June 19–24; New Orleans, LA. Piscataway, NJ: IEEE; 2022. p. 44–52.
Goel S, Pavlakos G, Rajasegaran J, Kanazawa A, Marik J. Humans in 4D: Reconstructing and tracking humans in transformers. Proceedings of the 2023 IEEE/CVF International Conference on Computer Vision (ICCV); 2023 October 2–6; Paris, France. Piscataway, NJ: IEEE; 2023. p. 14737–14748.
Agisoft. Metashape: Standard edition. https://www.agisoft.com/features/standard‐edition/. Accessed 31 Jan 2024.
Blender. Blender – A 3D modelling and rendering package. http://www.blender.org. Accessed 31 Jan 2024.
Meshlab. https://www.meshlab.net/. Accessed 31 Jan 2024.
CloudCompare. CloudCompare – 3D point cloud and mesh processing software Open Source Project. https://www.danielgm.net/cc/. Accessed 31 Jan 2024.
Devernay F, Faugeras O. Straight lines have to be straight. Mach Vis Appl. 2001;13:14–24. https://doi.org/10.1007/PL00013269.
Leica Geosystems. Leica FlexLine TS03 manual total station. https://leica‐geosystems.com/en‐us/products/total‐stations/manual‐total‐stations/leica‐flexline‐ts03. Accessed 31 Jan 2024.
Rajasegaran J, Pavlakos G, Kanazawa A, Malik J. Tracking people by predicting 3D appearance, location and pose. Proceedings of the IEEE Conference on Computer Vision and Pattern Recognition (CVPR); 2022 June 19–24; New Orleans, LA. Piscataway, NJ: IEEE; 2022. p. 2730–2739.
Bertillon A, McClaughry RW. Signaletic instructions including the theory and practice of anthro‐pometrical identification. Whitefish, MT: Kessinger Publishing; 1896.
Heuschkel ML, Labudde D. Reconsideration of bertillonage in the age of digitalisation: digital anthropometric patterns as a promising method for establishing identity. Forensic Sci Int Synergy. 2024;8:1004524. https://doi.org/10.1016/j.fsisyn.2023.100452.
Grant Information:
JP17K18379 Japan Society for the Promotion of Science; JP20K19835 Japan Society for the Promotion of Science; JP24K15003 Japan Society for the Promotion of Science
Contributed Indexing:
Keywords: 3D camera calibration; 3D human body shape model; SMPL‐X; forensic image/video analysis; height; person verification; real‐scale; silhouette; weight
Entry Date(s):
Date Created: 20250707 Date Completed: 20251104 Latest Revision: 20251104
Update Code:
20251104
DOI:
10.1111/1556-4029.70130
PMID:
40619616
Database:
MEDLINE
Weitere Informationen
The combination of computer vision (CV) and computer graphics (CG) is being developed for use in many fields. Consequently, reverse projection photogrammetry, which identifies geometric properties of a subject based on accurate reproduction of the image content, is beginning to replace analysis combining CV and CG. Since an image captured by a camera has two-dimensional (2D) geometry, estimating real-scale three dimensional (3D) information about a human or object from a low-resolution security camera image is a challenge and has not been achieved without prior knowledge of the person or object. However, deep learning technology that applies fitting a 3D human body shape model to a human image has been developed, but it is difficult to scale the reconstructed model to the actual scale with only a 2D image as an input. In this study, we propose a novel method to estimate a real-scale 3D human body shape model (SMPL-X model) from a human image via a combination of 3D camera calibration and CG-based reverse projection photogrammetry. The method estimates the position, orientation, posture, and body shape of a 3D human body shape model of a human image in a non-straight posture, which is difficult to analyze conventionally. The method was also used to estimate height and weight based on the estimated 3D human body shape, greatly expanding the range of analysis of height and weight estimation. The equal error rate from a few hundred to a few thousand comparisons was evaluated toward realizing person verification.
(© 2025 American Academy of Forensic Sciences.)