Welcome to Med Herb Lens

We are helping traditional medicine doctors and herbalists recognise and classify herbal plants faster

Get Started

About Us

A non-profit AI app that recognises herbal plants using images

  • Med Herb Lens is a non-profit project that utilises Artificial Intelligence (Deep Learning - Tensorflow) to assist medicinal herbalists and traditional healers and physicians in automatically recognizing and classifying medicinal plants.
  • This project was initiated and has been led by Mr. Minh Le, a Doctor of Traditional Medicine student from Thanh Dong University, Vietnam. He helped build the dataset and also assisted in evaluating the authenticity of the input data.
  • In consultation with Mr Le, this Android app has been developed by Mr Trien Tran as part of his Master of Information Technology degree\'s capstone thesis project. He is now an alumnus of St. Paul University Philippines. If you are an iOS developer or backend engineer with Nodejs knowledge and willing to join us voluntarily, please feel free to reach out.
  • As of now, the developed AI model can recognise only 70 herbal plants due to limited input data . Therefore, we look forward to expanding the number of recognised medicinal plants to almost 4000. This ambition will not be achieved without your support. There are 2 ways you may help us out:

    - Please contribute to the dataset by capturing your own herb images using your own mobile and camera devices, then upload them all to our server using this mobile app. The image quality does not matter, meaning that it is OK to set the lowest image resolutions for your camera. You may take photos in both portrait and horizontal modes, in different environments with different brightness (night vs day, indoor vs outdoor…). We need details such as fruits, blossoms, branches, leaves, roots, and whole plant, both fresh and sun-dried ones, if possible.

    - And can you please also help us fill in all the herbs' details such as usage, side effects, drug interactions etc.? To do this, please browse or search using our search bar in the mobile app to look for the herbs that you are familiar with or have information about. Then you can click on the edit buttons to modify those details at your own convenience.

The app's features

AI lens

You may look up a herbal plant via the Camera feature or using static images. There are 2 lookup modes: object detection and whole image modes.

Herbal Wiki

This app can also serves as a Wikipedia for herbal species. Please feel please to edit herb details at your convenience!

Upload herb images

Please contribute to our image dataset by uploading your own herbal photos. You can also view others' photos in the herb details screen.

Rating - Review

A fantastic feature allowing you to comment on the effectiveness and ease of use of the herbs you know. You can rate it too!

Label or classify herbal plants with AI lens

(Due to limited data collection, we currently support only 70 herb species. More coming soon!)

Imagine you can use your mobile device camera to tell a herb's name. You can even do more with a static single image or process multiple saved images at once.

We support 2 classification modes: via object detection (so that you can choose to label the detected herbal objects found in your images), or whole image mode, meaning that the AI lens will look at the entire image and give you the most probable result.

A Wikipedia for herbal plants

With approximately 4000 species, you are free to edit and modify herb details such as uses, side effects, precautions and more.

You are encouraged to capture your own herbal photos and upload them all to our server. Max 1000 photos allowed for each plant.

You may also rate the herbal medicines you have experience with on a scale of 5 and make a comment for each of them as well.

Research paper

This app was initially done as part of a master degree level Deep Learning research project. If you want to access our published research paper, please feel free to click the Research Paper button

Research Paper

Frequently Asked Questions

If you are unable to find answers to your questions here, please contact us directly via our contact details at the bottom of the page.

  • We currently support only 70 herbal plants. You can find the list of these by clicking the provided Link using the Link menu at the top of the page.

  • Well, we are workig toward training almost known 4000 herbal plants. But we don't have enough image data yet. You can find the full list by clicking the provided Link using the Link menu at the top of the page. These 4000 herbs are incorporated into our "Wikipedia" database, and thus you can look up any of them using the search bar in the app.

  • We are not sure yet because we have not had enough time and engineers for iOS and backend sides. If you know someone who would like to help out, please let us know! Thank you!

  • Please capture herb photos as many as you can and upload them all to our server using the upload feature in the herb detail screen in our app.

  • Nope! This app is non-profit and will never be commercialized. However, we may collect donations in the future if the maintenance cost becomes unaffordable to us.

  • We train the AI model using Tensorflow, a free and open-source software library for machine learning and artificial intelligence owned by Google LLC. Then, we deploy it to mobiles using ML Kit, an Android library for machine learning for mobile.

Our Team

Want to join our team voluntarily? Please send us an email!

Minh Le

Product Manager

Troy Tran

Fullstack Android engineer

To be partnered

iOS engineer

To be partnered

Backend engineer

Med Herb Lens

We welcome all your feedbacks. Please feel free to send us an email or reach out via social media such as Facebook page and Linkedin.

Adelaide, South Australia 5000

admin@medherblens.ml