LogMeal

Nutrition and food tracking have become increasingly important in today's health-conscious world, driving innovation in technology that helps people monitor their diets more effectively. From healthcare providers monitoring patient nutrition to athletes tracking their meal macros, there's a growing need for accurate and efficient food recognition systems. LogMeal has pioneered this space with an advanced API that identifies and quantifies over 1300 international dishes across multiple languages, offering sophisticated features for both casual users and professional applications. This technical introduction highlights how the company's decades of computer vision expertise has translated into practical solutions used by nutritionists, athletes, restaurants, and healthcare facilities worldwide.

Advanced Food Detection and Tracking

The LogMeal API offers advanced food recognition capabilities, supporting over 1300 international dishes across multiple languages, including English, Spanish, French, and German. Through sophisticated image analysis, the API distinguishes between prepared food, fresh produce, drinks, and sauces, while also identifying various food groups present in the image.

The API process involves multiple stages of image segmentation and analysis. When processing an image, the API first segments the image into distinct food items using either the standard segmentation endpoint or the quantity-estimating endpoint. This segmentation allows for accurate identification of each dish present in the image.

Following segmentation, the API employs advanced algorithms to recognize the specific dishes and estimate their quantities. For each identified food item, the system provides detailed information about the ingredients and their quantities, enabling precise nutritional analysis. This multi-step process allows for comprehensive food tracking and nutritional assessment, making it particularly valuable for applications in nutrition and healthcare.

API Features and Usage

The API provides robust support for both development and end-user applications through three primary endpoints:

1. Multiple Food Dishes Segmentation: This endpoint processes images to identify and recognize every food item within them. Successive development has refined this capability, as evidenced by daily image dataset updates that improve overall recognition accuracy.

2. Multiple Food Dishes Segmentation with Quantity Estimation: This advanced endpoint builds upon basic segmentation by not only identifying food items but also estimating serving sizes and quantities. This feature significantly enhances nutritional analysis capabilities.

3. POST /image/segmentation/complete/{model_version} and POST /image/segmentation/complete/quantity/{model_version}: These specific endpoint implementations enable developers to integrate advanced food recognition and tracking functionalities into their applications.

The API features a tiered pricing structure, offering 20 free images per day per user account. This generous free tier supports basic usage without requiring paid subscription initially. For more extensive usage, users can manage their plan through their account, with options to upgrade or downgrade subscription levels as needed.

Billing operates on a straightforward model based on base subscription pricing, with additional costs applied for image requests exceeding the included quota. Users have the flexibility to manage their accounts directly, including changing payment methods via debit or credit card information stored in their profiles.

Diverse Applications of LogMeal Technology

Nutritionists and Dietitians

The primary application of LogMeal's technology for nutrition professionals enables automated meal tracking and nutritional analysis. Registered users can upload photos of meals, and the API processes these images to identify and quantify all food items present. The system then generates detailed nutritional reports, which nutritionists and dietitians can use to monitor patient progress and adjust meal plans.

Athletes

For athletes, the technology provides personalized nutritional tracking across multiple diets and training regimens. Athletes can use their smartphone cameras to document their meals, and the API processes these images to provide accurate nutritional information. This feature supports both professional and amateur athletes in meeting specific dietary requirements for performance optimization.

Restaurants and Canteens

In the hospitality sector, LogMeal offers a self-checkout solution that combines food recognition with nutritional information. The system uses a dedicated scanner or integrated camera to process meals, automatically detecting and quantifying each dish. This technology supports both dine-in and takeaway orders, providing real-time nutritional data to customers while streamlining the ordering process for staff.

Healthcare Facilities

Patient nutrition tracking in healthcare settings benefits significantly from LogMeal's technology. In-house kitchens can use the system to monitor meal consumption, while external patients can use their smartphones to document their meals. The platform automatically generates nutritional reports, which healthcare providers can access to adjust treatment plans based on accurate dietary information.

Technical Implementation

The technology foundation for these applications relies on LogMeal's sophisticated image recognition processes. The system initially segments the image into distinct food items, identifying categories including prepared food, fresh produce, drinks, and sauces. Each food item is then processed through advanced algorithms to recognize specific dishes and estimate serving sizes. The platform supports multiple image processing endpoints, including basic segmentation and quantity estimation, allowing for flexible integration into various application frameworks.

Company Background and Expertise

LogMeal's technology foundation traces back over three decades of collective experience in computer vision between its founders PETIA (Research Advisor) and MARC (CTO), who established the company as CEO and Co-Founder. This robust technical expertise spans both industrial and academic domains, with the company having developed Deep Learning algorithms for seven years.

The company's research portfolio includes 25 international papers published on food recognition, underpinning their core capabilities. Their technological cornerstone is the world's largest food image dataset, continuously growing through daily image updates that refine food detection accuracy.

LogMeal partners with industry leaders such as Nestle, Porsche, and Unilever, while managing diverse client applications including food tracking for the elderly, medical solutions for diabetes and renal disease patients, autonomous checkout devices for self-service restaurants, and nutrition estimation for transplant patients.

The API's technical specifications enable 20 images per day per user account with premium plans scaling up to one image per second per user. Users manage their accounts directly, including changing payment methods via debit or credit card information stored in their profiles. The company offers flexible subscription models with custom plans available for alternative requirements, and users can upgrade or downgrade their plans at any time through their user profile.