How to efficiently send an image captured using react-native-vision-camera to my Express.js backend for processing?

I'm using react-native-vision-camera in my React Native app to capture images of plants. Once captured, I aim to send it to my backend (using Express.js and MongoDB) for identification.

Here's a snippet of my code where I capture the photo:

const capturePhoto = async () => {
  if (camera.current !== null) {
    const photo = await camera.current.takePhoto({});
    setImageSource(photo.path);
    setShowCamera(false);
    console.log(photo.path);
  }
};

Path is saved in imageSource state. After taking an image I show it on screen, and next to it there is a button Identify, when clicked handleIdentify function is triggered.

const handleIdentify = async () => {
    try {
      const base64String = await ImgToBase64.getBase64String(`file://${imageSource}`);
      const response = await axios.post(API_BASE_URL, {
        image: base64String
      }, {
        headers: {
          'Content-Type': 'application/json'
        }
      })
      console.log(response.data);
      console.log('base64: ',base64String)
    } catch(error) {
      console.error("Failed to identify plant:", error);
    }
  }

I've tried converting the image to Base64 format, but it seems the Base64 strings, especially for high-resolution images, can become quite large and cumbersome to send via API calls. I have gotten the 413 status code after trying this.
Now, my question this what would be the best way to solve this, and upload images to my backend? (Also I should note that this is for my University project, so my DB won't be populated with much data)

filestack has a free tier for about 500 image uploads.
And it has sdk for iOS and android. You can implement that to upload the high resolution images directly and send the public url to db. here is the link for the sdk's:
https://www.filestack.com/docs/uploads/pickers/

Update: you can upload high resolution images directly from your imagepicker generated uri using expo-file-system.uploadAsync to your preferred storage endpoint to generate public image url which then can be saved to db. This takes out the limit of 3mb that below steps required but also means the free tier storage will get used up fast.
File stack endpoint: "https://www.filestackapi.com/api/store/S3?key=MY_API_KEY"

I just encountered similar issues, but I have been working with expo-image-picker.
Here are the steps that worked for me for expo-image-picker. Used the default quality instead of 1.
Once image is picked, Add a check for the image size to be less than 3 mb using expo-file-system or something similar.
I sent base64 to backend and uploaded to (filestack or aws) to get a public link which I stored in mongodb.