OSAC Footwear & Tire Subcommittee Develops Process Map

An overview of the Footwear and Tire Examination Process Map developed by the OSAC Footwear & Tire Subcommittee

By Samantha Springer, a research assistant at the Center for Statistics and Applications in Forensic Evidence (CSAFE)

 

On June 8, 2022, the Organization of Scientific Area Committees for Forensic Science’s (OSAC) Footwear & Tire Subcommittee published a current practice document for footwear and tire examination.

The 37-page document consists of multiple process maps that cover a range of practices in the field of footwear and tire examination, including casts, gel lifts, known and unknown assessments, and different types of substrates with or without the presence of blood. Additionally, the flowcharts cover administrative processes such as verification and reporting, technical assessments, and administrative assessments.

The current practice document defines its purpose as five-fold:

  • help improve efficiencies while reducing errors,
  • highlight gaps where further research or standardization would be beneficial,
  • assist with training new examiners,
  • develop specific laboratory policies, and
  • identify best practices.

The document represents current practices instead of best practices and therefore does not necessarily endorse all the methodologies shown in the multiple process maps to ensure practitioners can find the process their lab uses. According to an article published by the National Institute of Standards and Technology (NIST), David Kanaris, Chair of the OSAC subcommittee, plans to release a more interactive version of the document in the future.

NIST facilitated the development of this process map through a collaboration between the NIST Forensic Science Program and OSAC’s Footwear & Tire Subcommittee.

Other OSAC subcommittees have released their process maps for other forensic science areas, including speaker recognition, DNA, friction ridge examinations, and firearms examinations.

CSAFE researchers Alicia Carriquiry, CSAFE director, and Jacqueline Speir, an associate professor at West Virginia University, are members of the OSAC Footwear & Tire Subcommittee.

Learn more about CSAFE’s work on footwear impression analysis at https://forensicstats.org/footwear/.

Insights: The Effect of Image Descriptors on the Performance of Classifiers of Footwear Outsole Image Pairs

INSIGHT

The Effect of Image Descriptors on the Performance of Classifiers of Footwear Outsole Image Pairs

OVERVIEW

Shoe prints left at a crime scene can often be partially observed, smudgy, or subject to background effects such as dirt or snow, which can make comparing prints to a reference image challenging. Similarly, prints from the same shoe can vary depending on the wearer’s gait, weight and activity during the time of impression. Reliable, qualitative methods have yet to be developed for visually assessing the similarity between impressions. To help develop such methods, researchers funded by CSAFE created an algorithm that extracts image descriptors (well-defined groups of pixels), then tested the algorithm by comparing simulated crime scene images to a study database.

Lead Researchers

Soyoung Park 
Alicia Carriquiry

Journal

Forensic Science International

Publication Date

February 2022

Publication Number

IN 128 FW

The Goals

1

Develop a quantitative method for comparing shoe print images.

2

Test this method’s performance against an existing “standard” method to quantify similarity between two images.

The Study

Park and Carriquiry created a study database of impression images, using 48 pairs of shoes which had been worn by volunteers for six months. They then scanned the shoe prints, placing 0 to 10 sheets of paper between the shoes and the scanner to simulate levels of degradation. In all, the researchers obtained 864 reference images, and made 1,728 pairs of images to compare
half of which were mated (coming from the same shoe), and half non-mated.

Meanwhile, the researchers developed an algorithm to compare these pairs using image descriptors, which identify distinct groups of pixels in an image such as corners, lines and blobs. In particular, they used the SURF and KAZE descriptors to identify blobs, and the ORB descriptor to identify corners.

A mated pair of images, scanned at level 0 and level 10 degradation

Using six different combinations of descriptors, the researchers ran their comparisons to determine which model had the best balance of accuracy and computation efficiency, which is required in real-world situations. For a control, they used a proposed method called Phase-Only
Correlation (POC) to compare to the descriptor-based methods.

SURF (Speeded-Up Robust Feature): a descriptor which uses a box filter on integral images

KAZE: meaning “wind” in Japanese, the name refers to the descriptor’s use of nonlinear diffusion filtering

ORB (Oriented FAST and Rotated BRIEF): a combination of two extraction methods, FAST (Features from Accelerated Segment Test) and BRIEF (Binary Robust Independent Elementary Features)

Results

Degradation Level 10

1

All tested models showed promise, with good quality images reaching accuracy of 95% or better, and even blurry images achieving accuracy of 85% to 88%.

2

The models that relied on the SURF and KAZE descriptors outperformed those that relied on ORB.

3

In comparison, the POC model failed to differentiate between mated and non-mated pairs.

Focus on the future

 

There is a lack of large databases with realistic footwear impressions. A larger database, with different brands and models of shoes, may help develop more robust algorithms for wider use.

Algorithms will likely never replace well-trained examiners, but the more accurate and efficient these algorithms become, the more useful they can be to examiners in their work.

Insights: An Algorithm to Compare Two-Dimensional Footwear Outsole Images Using Maximum Cliques and Speeded Up Robust Features

INSIGHT

An Algorithm to Compare Two-Dimensional Footwear Outsole Images Using Maximum Cliques and Speeded Up Robust Features

OVERVIEW

Footwear impression researchers sought to increase the accuracy and reliability of impression image matching. They developed and tested a statistical algorithm to quantify and score the degree of similarity between a questioned outsole impression and a reference impression obtained from either a suspect or a known database. The resulting algorithm proved to work well, even with partial and partial-quality images.

Lead Researchers

Soyoung Park 
Alicia Carriquiry

Journal

Statistical Analysis and Data Mining

Publication Date

21 Feb. 2020

Publication Number

IN 104 FW

The Goals

1

Develop a semi-automated approach that:

  • Compares impression evidence imagery with putative suspect or database images.
  • Calculates a score to quantify the degree of similarity (or correspondence) between the images.
  • Lowers human error and bias in current practice.

2

Create a method to obtain a similarity score for a pair of impressions which can be used to assess the probative value of the evidence.

APPROACH AND METHODOLOGY

This algorithm focuses on the similarity between two outsole images and relies on the concept of maximum clique. Local maximum cliques can be used to find corresponding positions in the two images so that they can be aligned.

Rotation and translation don’t affect a maximum clique –– it depends on the pairwise distances between nodes on the graph.

So –– although outsole pattern images may be translated, rotated and subjected to noise and other loss of information –– the geometrical
relationships between the points that constitute the pattern will not change much.

In this study, researchers developed a publicly available and usable database of 2D outsole impressions. Then the researchers used data from a KNM™ (knowledge navigator model) database.

Key Definitions

Graph Theory

Study of graphs made up of vertices connected by edges

Clique

A subset of vertices with edges linking symmetrically, where every two disinct vertices are adjacent

Maximum Clique

Clique that includes the larges possible number of vertices

KEY TAKEAWAYS FOR PRACTITIONERS

1

With this new comparison learning algorithm, practitioners can align images using features chosen as areas of interest and calculate a similarity score more objectively.

2

The proposed pattern-matching algorithm can work with partial images or images of variable quality by partially aligning patterns to quantify degrees of similarity between two impressions.

3

While this study focuses on footwear evidence, this algorithm has potential
applications for other situations of pattern comparison, like:

  •  latent prints
  • surveillance photos
  • handwriting
  • tire treads and more.

4

The algorithm can distinguish impressions made by different shoes –– even when shoes share class characteristics including degree of wear.

SEE THE ALGORITHM IN ACTION

Researcher Dr. Soyoung Park demonstrates the team’s novel algorithm in a CSAFE webinar. The method is promising, because it appears to correctly determine, with high probability, whether two images have a common or a different source, at least for the shoes on which they have experimented.

ShoeprintR

Explore and try the algorithm by downloading it.