Do image convolution by giving input matrix and filter matrix as sparse matries.
The first number describes the amount of matrix convolution we want to compute.
Following by each image matrix’s #row, #col, #non-zeros and its non-zero terms' row, col, value. Also, the filter matrix's #row, #col, #non-zeros and its non-zero terms' row, col, value.
Print out each output matrix’s #row, #col, #non-zeros and its non-zero terms' row, col, value.