Meal snapping apps are on the rise
Across the globe, meal snapping apps are gaining ground with consumers, and for good reason.
Image and video assisted meal tracking are becoming essential features in weight loss and food intake tracking apps, not to mention empowering dietitians and nutritionists with what they need to make fast, accurate dietary assessments.
Instead of keeping an unwieldy food diary or spreadsheet, simply upload a few photos or a record short video of your quinoa salad and allow your app to return an accurate list (or at least an estimation) of its nutritional contents - all complements of a human analyst and/or AI software, that learns and increases its success rate the more you use it.
It’s definitely a compelling use case. However, findings from various recent studies are not yet convinced of its long-term benefits. A common critique among researchers is that additional visual aid functionality within the app is not a substitute for evidence-based methods and strategies to help users actually lose weight or lower their risk of developing dietary-related illnesses (such as cardiovascular diseases and diabetes).
On the other hand, it’s still early days, especially for technology. The EU-funded goFOOD project is making use of high-powered AI to quickly estimate the nutritional content of meals to upend chronic and dietary-related illnesses. In one key study, it’s already performed better than experienced dietitians in visually recognising ingredients.
This is good news for personalised nutrition companies. Thanks to the increasing intelligence of AI, technology-assisted dietary assessment (TADA), along with time stamps and GPS data, will soon be even more accurate when diagnosing food-based visual media. Ultimately, this will mean less time spent logging and correcting food intake entries.
This makes the meal snapping feature massively valuable, and we foresee this playing right into the hands of personalised nutrition players. In this article we dive into this fascinating space of meal snapping apps otherwise known as food recognition apps.
The Meal snapping app market overview
The market for image recognition apps for nutrition is growing rapidly, as more people are becoming aware of the importance of nutrition for their health and well-being. According to a report by Grand View Research, the global market for food recognition and nutrition analysis was valued at USD 1.2 billion in 2020 and is expected to grow at a compound annual growth rate (CAGR) of 19.5% from 2021 to 20281
Drivers for the increased adoption of meal snapping apps
- The availability and affordability of smartphones and mobile devices, which enable users to access these apps anytime and anywhere.
- The advancement and innovation of computer vision and artificial intelligence technologies, which improve the accuracy and functionality of these apps.
- The emergence and popularity of social media platforms, which create a culture of sharing and comparing food photos among users.
- The growing awareness and interest of consumers in healthy eating, wellness, and fitness, which motivate them to monitor and improve their nutrition habits.
The Applications of Image Recognition Apps for Nutrition
Image recognition apps for nutrition have various applications and use cases, depending on the needs and goals of the users. Some of the common applications are:
- Calorie counting and nutrition tracking: These apps allow users to estimate the caloric value and nutritional content of their meals by simply taking a photo of their food. The apps then provide a breakdown of the macronutrients (carbohydrates, protein, fat, etc.) and micronutrients (vitamins, minerals, etc.) of the food, as well as the percentage of the recommended daily intake. Some of the apps also allow users to log their food intake and track their progress over time.
- Diet planning and coaching: These apps help users to create personalized diet plans and receive guidance and feedback based on their dietary goals, preferences, and health conditions. The apps can also suggest healthy alternatives, recipes, and tips to improve the user's nutrition choices. Some of the apps also incorporate gamification, rewards, and social features to motivate and engage the users.
- Food recognition and education: These apps enable users to learn more about the food they eat, such as the ingredients, origin, history, and culture. The apps can also provide information about the environmental and ethical impacts of the food production and consumption, such as the carbon footprint, water usage, animal welfare, and fair trade. Some of the apps also aim to raise awareness and promote action on food-related issues, such as food waste, food security, and food justice.
How the tech works
Image recognition apps for nutrition work by using computer vision and artificial intelligence to process and analyze the images of food captured by the user's camera. The basic steps involved in this process are:
- Image acquisition: The user takes a photo of their food using the app's camera or uploads an existing photo from their gallery. The app then crops, rotates, and adjusts the image to enhance its quality and clarity.
- Image segmentation: The app identifies and separates the different food items in the image, as well as the background, plate, and utensils. The app then assigns a bounding box or a mask to each food item to isolate it from the rest of the image.
- Image classification: The app labels each food item with its name and category, such as pizza, salad, fruit, etc. The app uses a pre-trained deep neural network, such as a convolutional neural network (CNN), to compare the features of the food item with a large database of food images and labels. The app then outputs the most probable label for the food item, along with a confidence score.
- Image analysis: The app estimates the portion size, volume, and weight of each food item, as well as the caloric value and nutritional content. The app uses various methods and techniques to measure the portion size, such as depth sensors, geometric models, reference objects, and user inputs. The app then multiplies the portion size by the caloric density and nutritional factors of the food item, which are obtained from a reliable nutrition database, such as the USDA National Nutrient Database. The app then outputs the results in a user-friendly format, such as a pie chart, a bar graph, or a table.
The Benefits of Image Recognition Apps for Nutrition
Image recognition apps for nutrition offer several benefits to the users, such as:
- Convenience and ease of use: These apps eliminate the need for manual entry, weighing, and measuring of food, which can be time-consuming, tedious, and inaccurate. Users can simply snap a photo of their food and get instant results, without having to search for the food name, select the portion size, or scan the barcode. These apps also work offline, which means users can use them even without an internet connection.
- Accuracy and reliability: These apps provide more accurate and reliable estimates of the caloric value and nutritional content of food than traditional methods, such as food labels, nutrition facts tables, and online calculators. These apps can recognize a wide variety of food items, including mixed dishes, homemade meals, and ethnic cuisines, which are often difficult to quantify and categorize. These apps can also account for the variations in the preparation, cooking, and serving of food, which can affect the nutrition quality and quantity.
- Awareness and education: These apps increase the user's awareness and knowledge of the food they eat, as well as the impact of their food choices on their health and well-being. These apps can help users to identify the nutritional gaps and excesses in their diet, as well as the sources of calories, macronutrients, and micronutrients. These apps can also provide users with useful information, tips, and recommendations to improve their nutrition habits and achieve their dietary goals.
Here are a few meal snappers flashing in nutrition and prevention apps
- Real-time insight into nutritional composition of meals and diet quality - just look at DietID for a glimpse into this space
- Feedback provided by apps and/or healthcare professionals based on food intake can be sharpened, further personalising (and humanising) the weight-loss and fitness arena. Noom is a good example of this happening right now
- An entire reseller market of new apps, food solutions and other monitoring technology can be opened up based on consumer data. For instance, consumers hoping to transition to a plant-based diet will have access to a range of retailers and online brands that closely align with their objective
- Eating out will likely also catch onto the trend. Those bold enough to invest in the tech to allow meal matching with restaurants which offer the same or similar dishes will reap the rewards. Suggestic is among those who are toying with this
The Opportunities and Challenges of Image Recognition Apps for Nutrition
Image recognition apps for nutrition have a lot of potential and opportunities to improve and expand their services and features, such as:
- Integrating with other devices and platforms, such as smart watches, fitness trackers, health apps, and social media, to provide a more comprehensive and holistic view of the user's health and wellness.
- Incorporating more advanced and innovative technologies, such as augmented reality, voice recognition, and natural language processing, to enhance the user experience and interaction.
- Developing more personalized and adaptive solutions, such as using the user's preferences, history, feedback, and biometric data, to tailor the app's functionality and content to the user's needs and goals.
- Collaborating with other stakeholders and partners, such as food producers, retailers, restaurants, health professionals, and researchers, to provide more accurate, reliable, and diverse data and information, as well as to create more value and impact for the users and the society.
Challenges and limitations of food recognition apps
- The complexity and variability of food, which can affect the app's performance and accuracy. Factors such as lighting, angle, distance, occlusion, distortion, and quality of the image can influence the app's ability to segment, classify, and analyze the food. Moreover, the diversity and ambiguity of food names, categories, and labels can pose difficulties for the app's recognition and interpretation of the food.
- The privacy and security of the user's data, which can be compromised or misused by the app or third parties. Users may have concerns about sharing their personal information, such as their food photos, dietary habits, health conditions, and biometric data, with the app or other users. Users may also be exposed to risks such as identity theft, fraud, or cyberattacks, if the app does not have adequate encryption, authentication, and authorization mechanisms.
- The ethical and social implications of the app's influence and intervention on the user's behavior and choices. Users may become dependent on or addicted to the app, which can affect their autonomy and agency. Users may also experience negative emotions, such as guilt, shame, or anxiety, if the app provides judgmental or unrealistic feedback or expectations. Users may also face discrimination or stigma, if the app reinforces or reproduces stereotypes or biases based on the user's food preferences, culture, or identity.
In a snap
Looking ahead, we anticipate massive adoption of image-based food tracking within the personalised nutrition world. We have no doubt that a picture will be worth a thousand opportunities, and more.
References
https://www.ncbi.nlm.nih.gov/pmc/articles/PMC7436102/
https://www.ncbi.nlm.nih.gov/pmc/articles/PMC5094328/
https://go-food.tech/
https://www.ncbi.nlm.nih.gov/pmc/articles/PMC7436102/
https://www.cambridge.org/core/journals/proceedings-of-the-nutrition-society/article/new-mobile-methods-for-dietary-assessment-review-of-imageassisted-and-imagebased-dietary-assessment-methods/33E6AE4398065ACEBACE8E9E3AD170C9#
Global Food Recognition and Nutrition Analysis Market Size Report, 2021-2028
Chen, X.; Johnson, E.; Kulkarni, A.; Ding, C.; Ranelli, N.; Chen, Y.; Xu, R. An Exploratory Approach to Deriving Nutrition Information of Restaurant Food from Crowdsourced Food Images: Case of Hartford. Nutrients 2021, 13, 4132. https://doi.org/10.3390/nu13114132
Amugongo, L.M.; Kriebitz, A.; Boch, A.; Lütge, C. Mobile Computer Vision-Based Applications for Food Recognition and Volume and Calorific Estimation: A Systematic Review. Healthcare 2023, 11, 59. https://doi.org/10.3390/healthcare11010059
Suggestic: https://www.suggestic.com