You should encode the image name before giving it to crawl for any bots, i.e. the image name should be as
%2F%D0%B7%D1%83%D1%80%D0%B0%D0%B3.jpg in your page source code.
All modern browser decode it and for users will show the UTF-8 characters in a human-friendly way.
So, do server-side encoding and encode all these characters before including in your source code.
View the source code for this Wikipedia page and you will understand:
Based on deeper review the website and other StackExchange users comments below. It looks that you must encode all of your URLs in your website and XML sitemap. In your website, you don't have encoded URLs, as well not encoded URLs in XML sitemap. Please use UTF-8 encoded URLs everywhere.
Keep in mind, that Bingbot is way behind Googlebot and just recently started to render JS based website in more wider scale. Using non-ASCII characters could cause bugs for Bingbot.
The URLs in your source code and DOM, as well in your XML sitemap must be encoded like this based on your example:
Here you can read more about proper sitemap encoding: https://www.sitemaps.org/protocol.html#escaping