Mode-Based Video Quantization for Image Quality and Bit-Rate Control
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Existing image processing technologies struggle to maintain consistent image quality for both static images and videos captured in different modes, leading to deterioration in image quality due to inappropriate quantization parameter settings, especially when switching between static image and video modes.
Innovation Solution
An image processing device that adjusts quantization parameters within specific ranges for different video capturing modes, allowing for optimized compression processing that maintains consistent image quality and controls bit rate, using MPEG encoding methods and considering factors like frame rate and scene discrimination.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Manufacturing precision
If the quantization parameter is decreased to improve image quality, then the image quality improves, but the bit rate increases and may exceed processing capacity
Solution Approach 1:
The patent applies dynamics by making the quantization parameter range adaptive rather than fixed. The compression processing section dynamically adjusts the quantization parameter range based on the video capturing mode (first or second mode) and the amount of generated code from past frames. This allows the system to optimize between image quality and bit rate control by selecting appropriate QP ranges from multiple predefined ranges, resolving the contradiction between maintaining high image quality and controlling bit rate within processing capacity.
2Quantity of substance
If the quantization parameter is increased to control bit rate, then the bit rate is controlled within processing capacity, but the image quality deteriorates
Solution Approach 1:
The system dynamically selects from multiple quantization parameter ranges based on video capturing mode and code amount feedback. Instead of using a single fixed QP value that would compromise image quality, the system has multiple predefined QP ranges and selects the appropriate range dynamically. This resolves the contradiction by allowing aggressive compression (higher QP) only when necessary while maintaining higher quality ranges when possible.
Solution Approach 2:
The patent changes the parameter range of the quantization parameter based on video capturing mode. Different video capturing modes (first mode for general video, second mode for static image extraction) have different optimal QP ranges. By changing the QP parameter range according to the mode and code amount, the system optimizes the balance between bit rate control and image quality, preventing excessive quality deterioration while controlling bit rate within processing capacity.
3Device complexity
If a single quantization parameter range is used for all video modes, then the device complexity is reduced, but the image quality consistency across different modes deteriorates
Solution Approach 1:
Rather than using a single fixed QP range for all modes, the system dynamically adapts the QP range based on the video capturing mode. The compression processing section receives mode information and selectively applies appropriate QP ranges from multiple predefined ranges. This dynamic adaptation ensures consistent image quality across different modes without requiring complex custom processing for each mode, as the same basic mechanism handles multiple modes with different parameters.
Solution Approach 2:
The patent implements parameter changes by defining multiple quantization parameter ranges corresponding to different video capturing modes. The first video capturing mode uses a first QP range, while the second video capturing mode (for static image extraction) uses a second QP range. This parameter differentiation ensures that each mode receives optimized compression settings, maintaining image quality consistency across modes while using a unified compression processing framework.
4Manufacturing precision
If the quantization parameter is set excessively small immediately after switching from static image to video mode, then the image quality is improved, but the bit rate exceeds processing capacity
Solution Approach 1:
The patent applies preliminary action by pre-defining multiple quantization parameter ranges for different video capturing modes before compression actually occurs. When switching from static image mode to video mode, the system immediately applies the appropriate QP range (second range for static image extraction mode) without needing to calculate or adjust parameters in real-time. This preliminary preparation prevents the problem of excessively small QP values by ensuring the correct QP range is selected from the start, balancing image quality and bit rate control.
Solution Approach 2:
The system changes the quantization parameter range based on the video capturing mode to prevent excessive bit rate generation. When in second video capturing mode (static image extraction), a specific QP range is applied that prevents QP from becoming excessively small, even when code amount is low. This parameter change approach ensures that image quality is maintained without allowing bit rate to exceed processing capacity, particularly during mode transitions.
Data Source
AI summary
Provided are an image processing device capable of performing compression processing on a video captured in each video capturing mode according to a first video capturing mode and a second video capturing mode with different capturing conditions. A possible range of a quantization parameter applied in a case where the video is compressed is made different in the first video capturing mode and the second video capturing mode with different capturing conditions. The quantization parameter is determined within a first range in a case of the first video capturing mode, and the quantization parameter is determined within a second range narrower than the first range in a case of the second video capturing mode. In particular, a second upper limit value of the second range is smaller than a first upper limit value of the first range, and a second lower limit value of the second range is larger than a first lower limit value of the first range.