Vectorize iterative addition in python arrays

For each element in a randomized array of 2D indices (with potential duplicates), I want to "+=1" to the corresponding grid in a 2D zero array. However, I don't know how to optimize the computation. Using the standard for loop, as shown here,

def interadd():
    U = 100 
    input = np.random.random(size=(5000,2)) * U
    idx = np.floor(input).astype(np.int) 

    grids = np.zeros((U,U))      
    for i in range(len(input)):
        grids[idx[i,0],idx[i,1]] += 1
    return grids

the runtime can quite significant:

>> timeit(interadd, number=5000)
43.69953393936157

Is there a way to vectorize this iterative process?