Would it be possible to do on-device OCR with a library like tesseract.js or another library?

General Q&A
1

We have the requirement to be able to do offline OCR. I have tried tesseract.js but could not get it to work with a image blob

import { createWorker } from 'tesseract.js';
function onPhotoChange(currentBlob, oldBlob, newBlob) {
    // photo is null when the photo is deleted on the capture-photo control
    if (currentBlob) {
        (async () => {
            try {
                const worker = await createWorker('eng');
                const ret = await worker.recognize(currentBlob.);
                console.log(ret.data.text);
                view.image_text = ret.data.text;
                await worker.terminate();
            } catch (error) {
                console.error('Text recognition failed:', error);
                view.image_text = 'Error: Failed to recognize text';
            }
        })();    
    }
}
2

@ajgreyling I am not sure, it will depend on the implementation of the specific JS library. I doubt you will be able to create workers directly in the runtime, but you may be able to use tesseract.js by including it in an HTML Bridge project or a TS HTML App Package