Introduce image management function for alpha image-matching code

Would suit skilled developers who like a challenge only. To introduce an image management function to the code @ [login to view URL] which we are using on a project - a pilot instance can be found at [login to view URL] At the moment, if we index 20,000 images and want to add 30,000 more, we have to reindex the full 50,000 images. Or indeed if we want to remove 5 images. Having the capability to add/remove batches of images would be an enormous help. However this requires someone who knows what they are doing. The code is very much ALPHA software, it does not come with much documentation, and it is not uncomplicated. There are cluster descriptions involved. Indexing files include the following files: [login to view URL], [login to view URL], [login to view URL], [login to view URL], [login to view URL], [login to view URL], [login to view URL], [login to view URL] There is a need to change the naming conventions of the software (perhaps using filenames) - at the moment, it allocates doc_id's to the line number of the filename in the [login to view URL] file. Initial Dockerfile attached, and I can provide indexing files created for a sample batch of 500 images.