Extract DEA Registration information from the document below. Some pages could not be read by OCR - they are marked "OCR-unreadable, see image below". Use the images (in order) for those pages.
Document content:
**==> picture [609 x 663] intentionally omitted <==**
**----- Start of picture text -----**<br>
��<br>
�<br>
��<br>� � �<br>� ��� <br>��<br>
�<br>
��<br>� � �<br>� ��� <br>**----- End of picture text -----**<br>
--- Page 2 (OCR-unreadable, see image below) ---
Pre-extracted NER entities:
- SCHEDULES: 3, 1
The images below correspond to the unreadable pages in order. Extract from both the markdown and the images.
Return ONLY a valid JSON object with: registrant_type, raw_registrant_name, provider_first, provider_last, company_name, dba_name, dea_registration_number, expiration_date, issue_date, schedules, business_activity, address_1, address_2, city, state, zip, zip_plus_4. Use null for missing optional fields.