To　effectively　monitor　the　development　of　crack　damage　in　precast　shear　walls　under　an　earthquake,　an　improved　probabilistic　imaging　analysis　method　based　on　the　piezoelectric　guided　wave　was　proposed.　The　detected　area　was　divided　into　multiple　sub-regions　by　arranging　array-sensors,　and　the　effective　signal　was　extracted　by　the　coherence　function　between　the　emitted　and　received　signals.　The　effective　signal　was　utilized　as　the　input　signal,　and　the　damage　indices-based　signal　energy　characteristic　difference　coefficient　were　calculated　for　each　sub-region.　Simultaneously,　the　damage　indexes　were　filtered　by　the　probability　and　statistical　method　to　eliminate　the　influence　of　environmental　factors.　The　damage　probability　imaging　algorithm　was　applied　to　superpose　imaging　for　all　the　cracks　in　the　detected　area　and　to　extract　the　image　threshold,　which　was　used　to　realize　the　judgment,　location　and　evolution　analysis　of　the　crack　damage　under　seismic　loads.　Subsequently,　the　effectiveness　of　the　proposed　method　and　the　crack　damage　evolution　was　verified　by　a　shaking　table　test　on　a　precast　frame-shear　wall　structure　in　this　paper.　The　results　show　that　the　proposed　damage　probability　imaging　method　can　directly　and　effectively　detect　the　occurrence,　location　and　evolution　of　crack　damage　under　an　earthquake.　It　provides　a　technical　guarantee　for　the　damage　monitoring　of　precast　structures.　©　2022,　Editorial　Office　of　Journal　of　Building　Structures.　All　right　reserved.