Ok, I can answer this because I went through this personally in the same field.
So my thoughts on patents vs. publications and future academic career in computer science:
Academic Currency: If you look at the university you want to attend, and then check the difference between a head of research group vs. a senior lecturer in the same research group, you will see, the publications play an important rule. People argue with me sometimes on this, but the buttom line is, you need to have a stack of publications to become a profesor.
Think About it the other way: Now in terms of intelectial properties, not all the universities have the staff to pay, so that they can draft the IP, and chase if for X amount of months to see the results. However, the publications keep the name of the university in conversations, as each author need to writes his/her affiliates; and it does not cost the university, because the academics/academics' students, who won their grants, will attend the conferences.
Get Ready To Be suprised!: From the quality of the papers to quality of softwre the academics produce, your mind will probably force you to go back to industry once in a while; because you will see lecturers who kissed programming languages goodbaye BUT they keep publishing papers and get promoted, because the game is not writting a solid backbone application for your database cluster, it is about producing research through publications.
What About Patents then? Well, there are academics I know personally, who mastered the patents' workflow by hiring other people, and they do produce a number of patents, once in a while. However, all of them are outside computer science field. It might help you to convince a good professor to take you for his/her research project, because you understand the logistics of uniquness and freshness of an idea. Also, it is also impressive to have it on your university profile, however at the moment at least, publications are still the main check point for each individual academic.