diff --git a/modelscope/pipelines/cv/ocr_detection_pipeline.py b/modelscope/pipelines/cv/ocr_detection_pipeline.py
index 62248714..b73f65a4 100644
--- a/modelscope/pipelines/cv/ocr_detection_pipeline.py
+++ b/modelscope/pipelines/cv/ocr_detection_pipeline.py
@@ -149,6 +149,8 @@ class OCRDetectionPipeline(Pipeline):
     def postprocess(self, inputs: Dict[str, Any]) -> Dict[str, Any]:
         rboxes = inputs['combined_rboxes'][0]
         count = inputs['combined_counts'][0]
+        if count == 0 or count < rboxes.shape[0]:
+            raise Exception('modelscope error: No text detected')
         rboxes = rboxes[:count, :]
 
         # convert rboxes to polygons and find its coordinates on the original image