Budget: 5000 UAH Deadline: 30 days
💡 Valuable essence: URL manipulation
The most effective way is to find "clean" originals. Often, websites store photos in folders like /cache/wm/ or add the suffix _watermark.jpg. Check the source code: if there is a path to the photo without the watermark, set up "Find and Replace" in SimplePars to substitute links using regular expressions. This will provide 100% quality without burdening the server.
⚙️ Technical solution: PHP processor + API
If there are no originals, use the "Image Processing" tab -> "Your PHP script." Write a script that will send each photo to an external API (for example, WatermarkRemover.ai), receive the cleaned file, and save it in OpenCart.
⚠️ Main risk: Performance and Budget
AI cleaning via API is expensive ($0.05–0.10 per photo) and slow (up to 5-10 seconds per image). When mass parsing 5,000+ products, this can cost hundreds of dollars and "crash" the parsing process due to server timeouts.
Question: Have you already checked the structure of links to the photos? Can you provide an example URL of a watermarked image so I can check the possibility of finding the original without AI?