Quantifying the similarity of 2D images using edge pixels: an application to the forensic comparison of footwear impressions

Park, Soyoung; Carriquiry, Alicia L.

doi:10.1080/02664763.2020.1779194

Cited by 3 publications

(3 citation statements)

References 24 publications

Supporting

Mentioning

Contrasting

Order By: Relevance

“…Hasegawa and Tabbone [16] used the Histogram of Radon Transform (HRT) as a descriptor for image features by decomposing the shape of the shoeprint image into its connected components. Park and Carriquiry [30] proposed novel similarity scores based on the differences and similarities between pairs of aligned images which were combined into a single value using a random forest. To hasten the calculation time, [30] rely on SURF [3] descriptors on which they implement their algorithm [29].…”

Section: Descriptor-based Methodsmentioning

confidence: 99%

See 1 more Smart Citation

A finely tuned deep transfer learning algorithm to compare outsole images

Jang

Park

Carriquiry

2023

Statistical Analysis

Self Cite

View full text Add to dashboard Cite

In forensic practice, evaluating shoeprint evidence is challenging because the differences between images of two different outsoles can be subtle. In this paper, we propose a deep transfer learning‐based matching algorithm called the Shoe‐MS algorithm that quantifies the similarity between two outsole images. The Shoe‐MS algorithm consists of a Siamese neural network for two input images followed by a transfer learning component to extract features from outsole impression images. The added layers are finely tuned using images of shoe soles. To test the performance of the method we propose, we use a study dataset that is both realistic and challenging. The pairs of images for which we know ground truth include (1) close non‐matches and (2) mock‐crime scene pairs. The Shoe‐MS algorithm performed well in terms of prediction accuracy and was able to determine the source of pairs of outsole images, even when comparisons were challenging. When using a score‐based likelihood ratio, the algorithm made the correct decision with high probability in a test of the hypothesis that images had a common source. An important advantage of the proposed approach is that pairs of images can be compared without alignment. In initial tests, Shoe‐MS exhibited better‐discriminating power than existing methods.

show abstract

Section: Descriptor-based Methodsmentioning

confidence: 99%

“…Park and Carriquiry [30] proposed novel similarity scores based on the differences and similarities between pairs of aligned images which were combined into a single value using a random forest. To hasten the calculation time, [30] rely on SURF [3] descriptors on which they implement their algorithm [29].…”

Section: Descriptor-based Methodsmentioning

confidence: 99%

A finely tuned deep transfer learning algorithm to compare outsole images

Jang

Park

Carriquiry

2023

Statistical Analysis

Self Cite

View full text Add to dashboard Cite

show abstract

“…In addition, the azimuth of object points a relative to the central eye β [19] of the active 3D panoramic vision sensor can be calculated.…”

Section: Asodvs Imaging Principle Analysis E Spatial Pointmentioning

confidence: 99%

Use of Digital Twins‐Based Intelligent Navigation Visual Sensing Technology in Environmental Art Design of Scenic Spots

Sun¹,

Zhou²

2022

Advances in Civil Engineering

View full text Add to dashboard Cite

The study expects to solve the bottleneck problem in the field of intelligent navigation visual sensing and build a network system of information physics, to achieve a visual sensor which can correspond the depth and color of a scene. It aims to improve the robustness and accuracy of color recognition of color structured light in a complex environment. Based on the digital twins (DTs) technology, the effective transformation of logistics process and physical entity to quasi-real-time digital image is realized. The omnidirectional vision sensing technology of a single viewpoint and the panoramic color volume structured light generation technology of a single emission point are integrated. The new active 3D panoramic vision sensor is achieved for which makes the current visual sensing technology developed into the visual perception of body structure. This technology is adopted in the preliminary design of environmental art in the scenic spot, and it can predict the design feasibility of environmental art in the scenic spot, avoid mistakes in decision-making to a great extent and save human and material resources. And also it can analyze and predict some possible dangerous situations, which can greatly improve the environmental safety factor of the scenic spot. In the improvement stage of environmental art design in the scenic spot, using active 3D vision sensing technology can obtain more comprehensive information and is more conducive to the selection of design schemes.

show abstract