Video Encoding Using Neighboring Block Information for Coding Efficiency

The patent applies preliminary action by pre-configuring the current block using information from neighboring blocks before the actual encoding/decoding process. The decoder pre-processes neighboring block data and prepares prediction information in advance, so that when the current block needs to be decoded, the necessary prediction data is already available, improving coding efficiency without significantly increasing processing complexity during the main decoding operation.

The patent uses copying by replicating the maximum coding depth configuration from neighboring LCUs to the current LCU. Instead of transmitting the maximum coding depth information for every LCU, the decoder copies this parameter from already-decoded neighboring LCUs. This maintains coding precision by ensuring consistent depth configuration while reducing information transmission overhead, as the copied parameter doesn't require additional bits in the bitstream.

The patent applies parameter changes by dynamically adjusting the reference picture candidate selection based on the cost information from lower depth CUs. When the cost indicates a good match, the system changes the parameter of reference picture selection to use fewer candidates, improving encoding speed. When the cost indicates a poor match, the system changes the parameter to consider more reference pictures, maintaining prediction accuracy. This adaptive parameter adjustment resolves the contradiction between encoding speed and prediction accuracy.