Educate engineering students in the field of multimedia, image codec and video codec

CUSTOMER'S TASK

Educate the Bachelor’s and Master’s engineering students in the field of Multimedia, image codec and video codec.

ELECARD SOLUTION

Elecard StreamEye Studio applications illustrate the pictorial coding concept used in video codecs to visualize the encoded bit stream and also verify the quality of the reconstructed videos.

RESULT

StreamEye Studio enables a thorough examination of each frame and each coding unit with various parameter and metrics for quality estimation can be carried out with ease and helps in comparing different codecs like H.264, H.265 etc. for better understanding of difference in encoding the video file.

IMPLEMENTATION

The following examples are a comparison between a 100-frame binary encoded with H.265 and a 504-frame RAW file in the .y4m extension. Elecard StreamEye Studio is used to obtain the different parameters such as Quantization parameter, coding unit size, prediction unit size, bit allocation, pixel intensities of each block and metrics such as Delta, SSIM, PSNR, MSE, APSNR, etc. (metric value for the Y, U and V can be obtained separately as well as the average value. These can also be analysed frame by frame instead of the entire stream).

Elecard StreamEye

• The graph above shows different parameters and metrics such as SSIM, bit allocation and QP for all the encoded frames.
• Partition, motion vectors, background of each coding block along with predictions can be observed.

Figure 1. Comparison between decoded bitstream and raw file

Elecard Video Quality Estimator

Video Quality Estimator allows users to compare binaries or bit streams of popular codecs like H.264 or H.265. Two binaries ‘A’ and ‘B’ can be compared with the RAW file as well as each other as seen in  Figure 2.

Figure 2. Comparison of HEVC bit stream with RAW file

Figure 3 illustrates the stream panel of Quest which shows the essential information about the stream such as stream type, profile, resolution, frame rate, quantization parameter values, the number of I or B slices and so on. It displays all the information that is needed to understand the stream that is selected. In this case, the bitstream selected had 100 frames, with a YUV ratio of 4:2:0 and a resolution of 1280x720. Users have a ton of information to work with.

Figure 3. Stream panel

Figure 4 illustrates the frame panel which displays information like the frame number, size, and quantization parameter values for current frame. This is displayed for the current frames of the two bitstreams A and B.

Figure 4. Frame panel

The graphics panel of quest is illustrated in Figures 5 and 6 in which graphs of selected files and metrics are used for comparison. (B-Slices are green, I-Slices are red and P-slices are blue in color).

Figure 5. Graphics Panel – PSNR metric

Figure 6. Graphics Panel – SSIM metric

Additional options such as clustered or overlapping, bar charts etc. can be selected.

The graph provides a visual interpretation of the metrics in order to compare bit streams with ease.

Elecard YUV Viewer

Figure 7 shows a frame by frame comparison of reconstructed file and the raw file along with pixel values of a chosen field.

Figure 7. YUV Viewer – Frame by frame comparison

Figure 8 shows a graphical representation of different metrics like PSNR, SSIM, MSE, NQI frame by frame for the reconstructed file and the raw file.

Figure 8. Graphics Panel – different metrics

Feedback

"Elecard tools for video analysis is a better way of understanding the video codec concepts with working on different encoder parameters and analysing the output video with respect to quality metrics."

Shreyanka Subbarayappa, Assistant Professor from Electronics and Communication Department