diff --git a/perception/autoware_crosswalk_traffic_light_estimator/README.md b/perception/autoware_crosswalk_traffic_light_estimator/README.md
index 514aeaf7813bc..0b4e2e818e9b7 100644
--- a/perception/autoware_crosswalk_traffic_light_estimator/README.md
+++ b/perception/autoware_crosswalk_traffic_light_estimator/README.md
@@ -104,7 +104,7 @@ end
 #### Update flashing flag
 
 <div align="center">
-  <img src="images/flashing_state.png" width=70%>
+  <img src="images/flashing_state.png" width=50%>
 </div>
 
 ## Assumptions / Known limits
diff --git a/perception/autoware_traffic_light_occlusion_predictor/README.md b/perception/autoware_traffic_light_occlusion_predictor/README.md
index 0a51de3dc2e3a..ab985c7b17c15 100644
--- a/perception/autoware_traffic_light_occlusion_predictor/README.md
+++ b/perception/autoware_traffic_light_occlusion_predictor/README.md
@@ -8,7 +8,7 @@ For each traffic light roi, hundreds of pixels would be selected and projected i
 
 ![image](images/occlusion.png)
 
-If no point cloud is received or all point clouds have very large stamp difference with the camera image, the occlusion ratio of each roi would be set as 0. The signal whose occlusion ratio is larger than max_occlusion_ratio will be set as unknown type.
+If no point cloud is received or all point clouds have very large stamp difference with the camera image, the occlusion ratio of each roi would be set as 0. The signal whose occlusion ratio is larger than `max_occlusion_ratio `will be set as unknown type.
 
 ## Input topics