18431287. Mode Dependent Intra Smoothing simplified abstract (Ofinno, LLC)
Contents
Mode Dependent Intra Smoothing
Organization Name
Inventor(s)
Alexey Konstantinovich Filippov of Khimki (RU)
Vasily Alexeevich Rufitskiy of Vladimir (RU)
Esmael Hejazi Dinan of McLean VA (US)
Mode Dependent Intra Smoothing - A simplified explanation of the abstract
This abstract first appeared for US patent application 18431287 titled 'Mode Dependent Intra Smoothing
Simplified Explanation
The patent application describes a method for projecting the location of a sample in a block to a point on a reference line using filter coefficients. Here is a simplified explanation of the patent:
- An encoder or decoder determines the location of a sample in a block and projects it to a point on a reference line.
- The encoder or decoder selects a filter coefficient table based on the angular direction and block size.
- First filter coefficients are selected from the table based on the fractional part of the displacement of the point on the reference line.
- The selected filter coefficients are applied to reference samples to determine a first filtered value.
- A prediction of the sample is then determined based on the first filtered value.
Potential Applications: This technology could be applied in image and video compression algorithms to improve the accuracy of predicting pixel values.
Problems Solved: This technology helps in reducing the error in predicting sample values, leading to better compression efficiency.
Benefits: Improved compression efficiency, higher quality images or videos, reduced data storage requirements.
Potential Commercial Applications: This technology could be utilized in multimedia applications, video streaming services, and digital image processing software.
Possible Prior Art: One possible prior art could be the use of filter coefficients in image compression algorithms to improve prediction accuracy.
Unanswered Questions: 1. How does the selection of filter coefficient tables impact the overall compression efficiency? 2. Are there any limitations to the size of the block that can be effectively processed using this method?
Original Abstract Submitted
An encoder or decoder projects, in an angular direction, a location of a sample in a block to a point on a reference line between a plurality of reference samples. The encoder or decoder selects a filter coefficient table, from among a plurality of filter coefficient tables, based on the angular direction and a size of the block. The encoder or decoder selects first filter coefficients from the filter coefficient table based on a fractional part of a displacement of the point on the reference line relative to the location of the sample. The encoder or decoder applies, to the plurality of reference samples, the first filter coefficients to determine a first filtered value. The encoder or decoder determines a prediction of the sample based on the first filtered value.
