Prediction method and device based on block form

The present invention relates to a prediction method and device based on a block form, in which a plurality of divided blocks are generated by dividing a block to be predicted when encoding and decoding a video. A prediction mode is derived for at least some of the plurality of divided blocks, and prediction may be performed on the plurality of divided blocks based on the derived prediction mode. When performing prediction for a divided block, information associated with a block to be predicted may be used, and information associated with other divided blocks that have been predicted prior to the divided block may be used. 本发明涉及一种基于块形式的预测方法和装置，其中，在对视频进行编码和解码时，通过对将被预测的块进行划分来产生多个划分块。针对所述多个划分块中的至少一些划分块引出预测模式，可基于引出的预测模式对所述多个划分块执行预测。在针对划分块执行预测时，可使用与将被预测的块相关联的信息，并且可使用与在划分块之前已被预测的其它划分块相关联的信息。