Image component prediction method, encoder, decoder, and storage medium
Patent Information
- Authority / Receiving Office
- JP · JP
- Patent Type
- Applications
- Current Assignee / Owner
- GUANGDONG OPPO MOBILE TELECOMMUNICATIONS CORP LTD
- Filing Date
- 2026-03-13
- Publication Date
- 2026-06-09
AI Technical Summary
【0012】 本出願の実施例は、画像成分予測方法、エンコーダー、デコーダー及び記憶媒体を提供する。現在のブロックの予測すべき画像成分の第1の基準ピクセルセットを決定し、第1の基準ピクセルセットから基準ピクセルサブセットを決定し、基準ピクセルサブセットには前記第1の基準ピクセルセットから選択された1つ又は複数の候補ピクセルが含まれ、基準ピクセルサブセットを用いて予測モデルのモデルパラメータを計算し、予測モデルが現在のブロックの予測すべき画像成分に対して成分間予測処理を行うために用いられ、このようにして、第1の基準ピクセルセットのスクリーニング処理により、重要でない基準ピクセル又は異常な基準ピクセルを除去することができるため、第1の基準ピクセルセット内のピクセルの数が減少され、計算の複雑さとメモリ帯域幅を低減することができるだけでなく、予測モデルの精度も向上させることができ、前記予測モデルが前記モデルパラメータによって前記予測すべき画像成分の予測処理を実現するために用いられるため、予測すべき画像成分の予測精度が向上し、ビデオ画像の予測効率が向上する。
Smart Images
Figure 2026094442000001_ABST
Abstract
Claims
1. A method for predicting image components applied to a decoder, The steps include determining a first reference sample set of the first image component of the current block, A step of determining model parameters of a prediction model based on the first reference sample set, wherein the prediction model is used to perform prediction processing on the first image component of the current block. An image component prediction method comprising the step of determining a predicted value of a first image component of the current block based on the model parameters of the prediction model and the reference block of the current block.
2. The step of determining a first reference sample set of the first image component of the current block is: The steps include determining the starting point position and sampling interval based on the number of reference samples adjacent to the top or left edge of the current block and a preset number of candidate samples, The step of determining a reference sample in the first reference sample set from a reference sample adjacent to the top or left edge of the current block, starting from the starting point position and according to the sampling interval. The method according to claim 1.
3. The step of determining the model parameters of the prediction model based on the first reference sample set is: The maximum value C of the reference sample in the first reference sample set. ｍａｘ and minimum value C ｍｉｎ The steps to determine, From a reference sample adjacent to the top or left edge of the reference block, the maximum value C of the reference samples in the first reference sample set. ｍａｘ and minimum value C ｍｉｎ The maximum value L of the adjacent sample reconstruction of the second image component corresponding to it. ｍａｘ and minimum value L ｍｉｎ The steps to determine, Said C ｍａｘ , C ｍｉｎ , L ｍａｘ and L ｍｉｎ Based on the above, a step of determining the model parameters of the prediction model, characterized by including The method according to claim 2.
4. The reference block of the current block is characterized in that it is the reconstruction block of the second image component of the current block. The method according to claim 1.
5. The aforementioned prediction model is characterized by being one of a linear model, a nonlinear model, or a multi-model. The method according to claim 1.
6. A method for predicting image components applied to an encoder, The steps include determining a first reference sample set of the first image component of the current block, A step of determining model parameters of a prediction model based on the first reference sample set, wherein the prediction model is used to perform prediction processing on the first image component of the current block. An image component prediction method comprising the step of determining a predicted value of a first image component of the current block based on the model parameters of the prediction model and the reference block of the current block.
7. The step of determining a first reference sample set of the first image component of the current block is: The steps include determining the starting point position and sampling interval based on the number of reference samples adjacent to the top or left edge of the current block and a preset number of candidate samples, The step of determining a reference sample in the first reference sample set from a reference sample adjacent to the top or left edge of the current block, starting from the starting point position and according to the sampling interval. The method according to claim 6.
8. The step of determining the model parameters of the prediction model based on the first reference sample set is: The maximum value C of the reference sample in the first reference sample set. ｍａｘ and minimum value C ｍｉｎ The steps to determine, From a reference sample adjacent to the top or left edge of the reference block, the maximum value C of the reference samples in the first reference sample set. ｍａｘ and minimum value C ｍｉｎ The maximum value L of the adjacent sample reconstruction of the second image component corresponding to it. ｍａｘ and minimum value L ｍｉｎ The steps to determine, Said C ｍａｘ , C ｍｉｎ , L ｍａｘ and L ｍｉｎ The features include the step of determining the model parameters of the prediction model based on the above. The method according to claim 7.
9. The reference block of the current block is characterized in that it is the reconstruction block of the second image component of the current block. The method according to claim 6.
10. A computer storage medium storing an image component prediction program, wherein when the image component prediction program is executed by a processor, the computer storage medium implements the following image component prediction method to generate a bitstream: The steps include determining a first reference sample set of the first image component of the current block, A step of determining model parameters of a prediction model based on the first reference sample set, wherein the prediction model is used to perform prediction processing on the first image component of the current block. A step of determining the predicted value of the first image component of the current block based on the model parameters of the prediction model and the reference block of the current block.