<p>In the real world, the sparsity of signal is always unknown, making it challenging to determine the appropriate iteration stopping criteria and measurement matrix dimensions for compressed sensing algorithms. This uncertainty often leads to degraded compression and sensing performance. To address this issue, we propose a stopping criterion based on the matching ratio of binary channel occupancy status. Initially, the measurement matrix size is set by a conservative value, which is then gradually increased to improve the matching ratio. When the stopping criterion is met, the suitable number of compressed measurements is obtained, thereby avoiding the problem of excessively large measurement sizes. Additionally, an incremental orthogonal matching pursuit algorithm is introduced, which reduces computational complexity by reusing previous calculation results. Experimental results demonstrate that the proposed algorithm not only achieves a favorable compression ratio but also performs well in spectrum sensing, making it applicable to real-world signals and possessing strong practical value.</p>