DECEMBER 1981 | JASWANT R. JAIN, MEMBER, IEEE, AND ANIL K. JAIN
A new technique for estimating interframe displacement of small blocks with minimum mean square error is presented. The method involves an efficient 2D search procedure to determine the direction of displacement. The technique was applied to two image sets, resulting in 8–10 dB improvement in interframe variance reduction due to motion compensation. Motion compensation is used in designing a hybrid coding scheme, achieving a factor of two gain at low bit rates.
The paper discusses displacement measurement and its application in interframe image coding. It describes a method for measuring displacement between frames by dividing images into smaller blocks and estimating the direction of minimum distortion. The algorithm uses a 2D search method to find the displacement vector with high accuracy. The results show that motion compensation significantly reduces the variance of interframe differences, improving the signal-to-noise ratio (SNR) and enabling better compression.
The technique was tested on two image sets, "Cronkite" and "Chemical Plant," each containing 16 frames of 256×256 pixels. The results demonstrated that motion compensation reduces interframe variance, leading to improved compression performance. The method was also applied to data compression, showing that motion compensation can significantly improve the quality of compressed images.
The paper also discusses the application of motion compensation in data compression, particularly in frame skipping and interpolation. It shows that motion compensation can reduce the bit rate by up to a factor of two, making it suitable for low-bit-rate applications such as videotelephone and video conferencing. The results indicate that motion compensation improves the performance of interframe predictive coders, especially when the temporal correlation is not very high.
The paper concludes that the proposed displacement measurement technique is effective for interframe coding and can significantly improve the performance of image compression systems. The method is based on dividing images into blocks, estimating displacement vectors, and using these vectors for motion compensation in hybrid coding schemes. The results show that motion compensation can reduce the bit rate by up to a factor of two, making it a valuable tool for image compression.A new technique for estimating interframe displacement of small blocks with minimum mean square error is presented. The method involves an efficient 2D search procedure to determine the direction of displacement. The technique was applied to two image sets, resulting in 8–10 dB improvement in interframe variance reduction due to motion compensation. Motion compensation is used in designing a hybrid coding scheme, achieving a factor of two gain at low bit rates.
The paper discusses displacement measurement and its application in interframe image coding. It describes a method for measuring displacement between frames by dividing images into smaller blocks and estimating the direction of minimum distortion. The algorithm uses a 2D search method to find the displacement vector with high accuracy. The results show that motion compensation significantly reduces the variance of interframe differences, improving the signal-to-noise ratio (SNR) and enabling better compression.
The technique was tested on two image sets, "Cronkite" and "Chemical Plant," each containing 16 frames of 256×256 pixels. The results demonstrated that motion compensation reduces interframe variance, leading to improved compression performance. The method was also applied to data compression, showing that motion compensation can significantly improve the quality of compressed images.
The paper also discusses the application of motion compensation in data compression, particularly in frame skipping and interpolation. It shows that motion compensation can reduce the bit rate by up to a factor of two, making it suitable for low-bit-rate applications such as videotelephone and video conferencing. The results indicate that motion compensation improves the performance of interframe predictive coders, especially when the temporal correlation is not very high.
The paper concludes that the proposed displacement measurement technique is effective for interframe coding and can significantly improve the performance of image compression systems. The method is based on dividing images into blocks, estimating displacement vectors, and using these vectors for motion compensation in hybrid coding schemes. The results show that motion compensation can reduce the bit rate by up to a factor of two, making it a valuable tool for image compression.