You'll need a better approach.
Please post the HTML in question, not just a link to the site. The link is helpful but we'd like the code here too please. The site will go away or change and that would make the question hard to understand.
You should not use the _1UoZIX identifier. It's generated on the fly (dynamically) and changes over time. It also has no semantic meaning, doesn't say fridge at all
To get to that specific image
alt="Godrej 260 L Frost Free Double Door Refrigerator"
is an attribute I would consider basing my locator on in comparison to the layout and react id's. Even that is text based and not a great selector but the page in question is not well constructed for other tag identifiers for this image as far as I can see. You could use *fridge* perhaps on the alt tag.
However, getting to the 'first search result image' element generically, i.e. first result and not just when it is that fridge... is going to be a challenge given the layout. One approach that might work is (pseudo code) a locator that is something like
//div a[href='/']../img[2]
i.e. find the main div based on the home anchor within it, go back up one level (xpath relative addressig) and then look for the second image within that div. I don't like this much but this is due to the page structure presented. Some version of this may work. You'll need to read up on xpath relative addressing if you don't know it already.
fyi The ability to transverse relative paths is the only reason I will use an XPATH locator over CSS and so it is very rare in my experience.