Current Research
OUR CURRENT RESEARCH continues to investigate improved fit, ranging from evaluating protective clothing for firefighters to measuring how clothing moves over the body in different positions.

Design of Firefighter Uniforms

This study investigated the comfort and effectiveness related to the design of firefighter uniforms. First we conducted a focus group meeting with 6 firefighters and the fire chief in Ithaca, NY. This provided us with information on the problems that firefighters have with their gear, including requests for: less bulky and lighter gear; fewer restrictions in the crotch of the pants; less restriction from the air tank; more comfortable suspenders; more dexterity in the gloves; and adjustable fit for weight loss and gain. We then identified 5 positions, including three active poses and two range of motion postures, that would help us isolate potential areas of improved fit.

Three firefighters were scanned in 6 poses (a standing pose, 3 active poses, and 2 range of movement poses) for 6 clothing conditions (station uniform and 5 turnout gear models). Preliminary work on the scans shows good potential for analysis of fit, using merged scans of the firefighter in station gear in each position, and scans of the same firefighter in the suits.

When the firefighter's suit scan is translucent, you can see the relationship of the body to the suit in three dimensions. (Image: Cornell Body Scan Research Group)
Firefighters' protective gear is both heavy and bulky. (Image: Cornell Body Scan Research Group)

Bra and Body Shaper Research

This research project is designed to investigate bra and body-shaper sizing, fit, and styling based on standard sizing techniques. First, we will determine if the subject is wearing the correct bra size, based on her measurements. Then the subject will try our bra and body shaper styles in the appropriate size, and we will use our body scanner to create 3D images of these garments. In addition, we will use pressure sensors to measure the clothing pressure at key areas. We will also gather information on the subject's reactions to the size, fit, and style of these products.

Front and side views of a subject wearing five different models of bra. (Image: Cornell Body Scan Research Group)
Bodyshapers come in many styles. (Image: Cornell Body Scan Research Group)

Virtual Fit Avatars

In this study, we compared reliability, acceptability to the consumer, and predictive validity of consumers’ responses to three kinds of virtual garment fit images: from the internet site My Virtual Model™ (MVM); an apparel CAD program ‘i-designer’; and the 3D body scanner.

We made three types of virtual models of 30 female participants, ages 18 to 25, who fell within the selected jean sizes 6 to14. The test garment chosen for the study was a boot cut blue jeans style with slight flare since it was likely to provide the most acceptable fit within the participant pool and is a preferred style of the selected target market. The participants were scanned with our Human Solutions Vitus body scanner for measurements. They then tried on a pair of the study jeans and were scanned and photographed in the jeans.

A scanned pair of denim jeans, which can be overlaid on a subject's avatar. (Image: Cornell Body Scan Research Group)

The first avatar was created using the measurements from the body scan according to the instructions on the MVM™ website. A second avatar was created in the patternmaking and virtual fit software ‘i-designer’, which uses a computer algorithm to construct the avatar from the 3D body scan data. A pattern was made from the test jeans, and these jeans were virtually fitted to each participant’s avatar in ‘i-designer’. Finally, a third avatar was made from each participant’s body scan. The scan was modified to smooth areas of the body and to compress the stomach, as the jeans do when worn. A separate scan of the jeans was overlaid on the body to demonstrate fit.

Participants returned to view their three avatars and rated the avatars on how accurate they felt each avatar was, which image they liked the best, and which image they would trust as a virtual fit avatar. They were also asked if the virtual fit experience would make them more likely to purchase online, and how confident they would be that they would find good fit based on each different virtual avatar.

The scanner software can calculate the body measurements needed to create the avatars. (Image: Cornell Body Scan Research Group)

Sizing Systems for Athletic Apparel

This research study is a collaborative effort between Nike and Cornell University, conducted by the apparel class FSAD 664: Anthropometrics and Apparel. Nike is interested in testing their apparel sizing, and this study seeks to determine how well their running shorts and t-shirts fit on a variety of study participants.

Nike has provided us with T-shirts and running shorts for the study. We have scanned the participants wearing the garments and asked them to fill out surveys recording their responses to the garments’ fit. A subset of 10 participants were also given two pairs of shorts to test in a wear test for one week. They recorded their impressions on fit and function in a journal format. The students who conducted this study learned the difficulties associated with assessing fit, and comparing expert fit analysis to the results of participant's fit assessments.

Front and back images of a participant in the running shorts and t-shirts. (Image: Cornell Body Scan Research Group)
The surveys gathered information on wear, fit and buying preferences. (Image: Cornell Body Scan Research Group)

Lower Body Surface Change

Previous research on body measurements by 3-D body scanners has mostly been based on a standing position. A seated position is, of course, used frequently in daily life, and therefore, this pose would be significant in future research of body measurements by 3-D body scanners. The subjects of this research were 25 female university students of average physical shape. Body surface length and circumference of lower body were measured when the subjects took four positions: standing; 135-degree knee-bend; one step forward; and sitting with 90-degree knee angle. By examining the differences of body surface length and circumference in each posture and calculating appropriate ease of circumference and length in pants patterns that affect wearer's comfort and movement, this study intended to provide data necessary for the development of pants patterns.

Subject shown in same pose, minimally clothed and with pants. The measurement points and lines are marked. (Image: Cornell Body Scan Research Group)
Here, the seated measurements are shown. (Image: Cornell Body Scan Research Group)

Fabric Drape Modeling

We have made sets of skirts in a straight and a flared pattern in five different fabrics which vary in fiber content, weave structure and weight. We then scan each garment on a dress form and also photograph it. In the future, Steve Marschner from Cornell's Computer Science Department and his group will use the scan data and digital versions of the patterns to analyze the mathematics of fabric drape.

Scans of flared skirts (denim on the left and lightweight cotton on the right) rotate, showing the differences in drape. (Image: Cornell Body Scan Research Group)
The same skirt is constructed in different fabrics. On the left is a denim model and on the right, a lightweight cotton. (Image: Cornell Body Scan Research Group)

Virtual Fit of Digital Garments on 3D Body Scans

3D body scanners can be used to create realistic images of a person, which can then be used as avatars. By importing these avatars into a patternmaking software which has 3D virtual try on feature, it is possible to rotate 3D images to see how the garment would look on a person. Depending on the fit and style preferences, the garment pattern could be also altered by the users themselves if they have access to the software or if an interface can be created to modify patterns.

This study incorporates 3D body scanning and Optitex PDS CAD with 3D Virtual Fit application capabilities into the process of virtual try-on. Eight study participants, who were working with these technologies as a part of an independent study class, used data from our Human Solutions 3D body scanner to create and import their virtual images into Optitex CAD patternmaking system. They then designed and created patterns for garments which they viewed in a virtual image on their avatars.

In addition to providing well fitted fashion pieces with the help of 3D body scanner data and virtual try-on, this process can also empower consumers with design authorship. In future greater engagement in the process of getting clothes that we value can change our relationship to throw-away fashion, resulting in less waste. Currently, we are examining the data to understand the potentials of these digital technologies to convert consumers into designers of their own products as well as the underlying factors regarding the design authorship-sustainable consumption interaction in more detail.

Click on the image to see the digital garment and avatar in motion. (Image: Cornell Body Scan Research Group)
A subject wearing the garment which she designed for herself. (Image: Cornell Body Scan Research Group)

Categorization of Lower Body Shapes

Previous research on lower body shape analysis is mostly based on subjective comparison of body silhouettes or by calculating ratios of waist and hip girths. However, our study investigated a reliable and objective way of analyzing lower body shapes and postures using statistical analysis based on 3D scan measurement data. Analysis of 2,488 women aged 18 to 35 in SizeUSA (national sizing survey) data revealed five components that can represent a distinctive lower body shape:

  1. waist to top hip silhouette [PC1]
  2. top hip to full hip silhouette [PC2]
  3. buttocks prominence [PC3]
  4. abdomen prominence [z-score 1]
  5. slope from abdomen point to front hip point [z-score 2]
Based on these five components, body shapes can be categorized into three types: curvy shape, hip tilt shape, and straight shape.

After the analysis, we recruited 83 participants in Ithaca, NY, and scanned their bodies with our Human Solutions Vitus 3D scanner. The scan files were then measured using the same [TC]2 automatic measuring software used for the SizeUSA survey. Each participant’s body shape group and her position within the group were identified based on her measurements. Then, the best three representative ‘fit models’ were selected from within each group.

The five components representing lower body shape. (Image: Cornell Body Scan Research Group)
The three fit models. (Image: Cornell Body Scan Research Group)

Fit of Jeans and Body Satisfaction

In this study we measured the degree of satisfaction with the body related to clothing fit using a body cathexis scale. High ratings on a body cathexis scale are related to high confidence levels. Clothing fit is an important factor in the management of clothing- mitigated body appearance and, potentially, body satisfaction. We wished to determine whether the college-age female wearer’s perception of denim jeans fit is related to her degree of body satisfaction.

For this study 313 undergraduate female students completed an online survey about their denim jeans wardrobe and preferences in terms of fit, style and cost. A subset of 36 respondents brought in their worst fitting and best fitting jeans, modeled the jeans in front of a mirror, discussed the fit of each pair, and were body scanned in minimal clothing (underwear, tank top and leggings) and in each pair of jeans. Participants were surveyed on the fit, style, and cost of the jeans as well as their feelings of security and comfort. Participants completed a lower body cathexis scale when they first entered the study room, after they were scanned in minimal clothing, and after they were scanned in each pair of jeans. Overall Cathexis scores were lowest in the poorly fitting jeans, next lowest in minimal clothing, and highest in the well fitting jeans. Participants felt significantly more secure and more comfortable in the well fitting jeans. Follow-up studies are planned on perception of fit based on ease values, and cross cultural issues in body satisfaction and fit.

This graph shows initial body satisfaction averages (street clothes) compared to the scan uniform and the two pairs of jeans for each lower body part. (Chart: Cornell Body Scan Research Group)
Each study participant was scanned in a pair of poorly fitting jeans (A) and a pair of well fitting jeans (B) from her wardrobe. (Image: Cornell Body Scan Research Group)