Image-based food volume estimation

Xu, Chang; He, Ye; Khanna, Nitin; Parra, Albert; Boushey, Carol J.; Delp, Edward J.

doi:10.1145/2506023.2506037

Cited by 55 publications

(32 citation statements)

References 26 publications

Supporting

Mentioning

Contrasting

Unclassified

Order By: Relevance

“…We denote the 3D graphical model that is reconstructed from multiple-views as a pre-built 3D model [11]. In addition to pre-built 3D models we have added pre-defined 3D models for conventional shapes [12]. Using the camera parameters we can project both the pre-built and pre-defined 3D models of each food item back onto the image plane then the food volume can be estimated based on a similarity measure of the back-projected region overlaid on the food image segmentation mask.…”

Section: Introductionmentioning

confidence: 99%

See 1 more Smart Citation

Single-View Food Portion Estimation Based on Geometric Models

Fang

Liu

Zhu

et al. 2015

2015 IEEE International Symposium on Multimedia (ISM)

View full text Add to dashboard Cite

In this paper we present a food portion estimation technique based on a single-view food image used for the estimation of the amount of energy (in kilocalories) consumed at a meal. Unlike previous methods we have developed, the new technique is capable of estimating food portion without manual tuning of parameters. Although single-view 3D scene reconstruction is in general an ill-posed problem, the use of geometric models such as the shape of a container can help to partially recover 3D parameters of food items in the scene. Based on the estimated 3D parameters of each food item and a reference object in the scene, the volume of each food item in the image can be determined. The weight of each food can then be estimated using the density of the food item. We were able to achieve an error of less than 6% for energy estimation of an image of a meal assuming accurate segmentation and food classification.

show abstract

Section: Introductionmentioning

confidence: 99%

“…We have also examined the use of prism models (an area-based volume model) that either have non-rigid shapes or do not have significant 3D structures (e.g. scrambled eggs) [12], [13]. Our previous portion estimation technique requires manual initialization of the parameters for different food types prior to use [5], [12].…”

Section: Introductionmentioning

confidence: 99%

Single-View Food Portion Estimation Based on Geometric Models

Fang

Liu

Zhu

et al. 2015

2015 IEEE International Symposium on Multimedia (ISM)

View full text Add to dashboard Cite

show abstract

“…Subsequently, the mobile food record (mFR) was modified and adapted for field use with the CoASTAL cohort. The application, running on the Technology Assisted Dietary Assessment (TADA) system for mobile devise use, assessed dietary intake via participant captured images of food and beverages they were about to consume (Boushey et al, 2009; Zhu et al, 2010; Xu et al, 2012; He et al, 2013a; He et al, 2013b; Xu et al, 2013a; Xu et al, 2013b; Wang et al 2015; Zhu et al, 2015). In addition to analyzing images of razor clam meals, the application was also designed to capture the source beaches for the consumed razor clams, and to collect real-time memory data.…”

Section: Methodsmentioning

confidence: 99%

“…The fiducial marker is about 2-inches square. This item is included in every image as a color and size reference to help with the reconstruction of a three-dimensional environment that allows for estimation of the volume of the foods and beverages (Lee et al, 2012; Xu et al, 2012; Xu et al, 2013b). In this study, the food items were calibrated to represent a variety of razor clam preparations, including clam chowder.…”

Section: Methodsmentioning

confidence: 99%

Dietary assessment of domoic acid exposure: What can be learned from traditional methods and new applications for a technology assisted device

et al. 2016

View full text Add to dashboard Cite

Three Tribal Nations in the Pacific Northwest United States comprise the members of the CoASTAL cohort. These populations may be at risk for neurobehavioral impairment, i.e., amnesic shellfish poisoning, from shellfish consumption as a result of repeated, low-level domoic acid (DA) exposure present in local clams. Previous work with this cohort confirmed a high proportion of clam consumers with varying levels of potential exposure over time. Since clams are an episodically consumed food, traditional dietary records do not fully capture exposure. Frequency questionnaires can capture accumulated doses over time and this data can be used to examine dose-response relationships with periodic studies of memory and learning. However, frequency questionnaires cannot be used to assess consumption and memory response in real time. To address this shortcoming, a modified technology assisted dietary assessment (TADA) iPod application was developed to capture images of the clam meal, sourcing data, and associated memory functioning within 24 hours and seven days after consumption. This methodology was piloted with razor clam meals consumed by members from the CoASTAL cohort. Preliminary findings suggest that the TADA iPod application is potentially useful in collecting real-time data with respect to razor clam consumption, as well as one day and seven day memory outcome data. This technology holds promise for addressing the challenges of other HAB related dietary exposure outcome studies.

show abstract

“…Martin et al assume that a correlation exists between the area of the region where food is visible in an image and its actual volume [6]. To achieve the same goal, the system described in [7] first detects the camera pose using a fiducial marker placed within the scene and then tries to match the food item by generating several pre-defined 3D shape models of which the best-matching one is adopted. However, this approach works only for the items that can be approximated by the regular shapes it takes into consideration.…”

Section: Introductionmentioning

confidence: 99%

A New Approach to Image-Based Estimation of Food Volume

et al. 2017

View full text Add to dashboard Cite

Abstract:A balanced diet is the key to a healthy lifestyle and is crucial for preventing or dealing with many chronic diseases such as diabetes and obesity. Therefore, monitoring diet can be an effective way of improving people's health. However, manual reporting of food intake has been shown to be inaccurate and often impractical. This paper presents a new approach to food intake quantity estimation using image-based modeling. The modeling method consists of three steps: firstly, a short video of the food is taken by the user's smartphone. From such a video, six frames are selected based on the pictures' viewpoints as determined by the smartphone's orientation sensors. Secondly, the user marks one of the frames to seed an interactive segmentation algorithm. Segmentation is based on a Gaussian Mixture Model alongside the graph-cut algorithm. Finally, a customized image-based modeling algorithm generates a point-cloud to model the food. At the same time, a stochastic object-detection method locates a checkerboard used as size/ground reference. The modeling algorithm is optimized such that the use of six input images still results in an acceptable computation cost. In our evaluation procedure, we achieved an average accuracy of 92% on a test set that includes images of different kinds of pasta and bread, with an average processing time of about 23 s.

show abstract

Image-based food volume estimation

Cited by 55 publications

References 26 publications

Single-View Food Portion Estimation Based on Geometric Models

Single-View Food Portion Estimation Based on Geometric Models

Dietary assessment of domoic acid exposure: What can be learned from traditional methods and new applications for a technology assisted device

A New Approach to Image-Based Estimation of Food Volume

Contact Info

Product

Resources

About