Use encoding reported by camera

The camera driver reports its encoding in the sensor_msgs::Image. This
string is taken from include/sensor_msgs/image_encodings.h. The most
common ones:
    "mono8"
    "bgr8"
    "bgra8"
    "rgb8"
    "rgba8"
    "mono16"
are also used for cv_bridge::toCvCopy. So it can be simply passed on.

darknet_ros/src/YoloObjectDetector.cpp

@@ -162,7 +162,7 @@ void YoloObjectDetector::cameraCallback(const sensor_msgs::ImageConstPtr& msg) {
   cv_bridge::CvImagePtr cam_image;
 
   try {
-    cam_image = cv_bridge::toCvCopy(msg, sensor_msgs::image_encodings::BGR8);
+    cam_image = cv_bridge::toCvCopy(msg, msg->encoding);
   } catch (cv_bridge::Exception& e) {
     ROS_ERROR("cv_bridge exception: %s", e.what());
     return;