Overview
Menu

Media Quality Inspection Template

Last updated: 2024-09-14 10:58:40Download PDF

Overview

The media quality inspection service can comprehensively identify issues in media assets, verify whether the contents meet the requirements, and offer suggestions. Users may directly use the preset templates or create media quality inspection templates in the console to customize quality inspection parameters.

The quality inspection contents include:
Format diagnosis: Media diagnosis is a cloud-based audio and video processing service designed for real-time diagnosis of media streams. It can be used to analyze exceptions in stream information, temporal information, stream status, container encapsulation, and decoding.
No-reference scoring: It refers to scoring the quality of a video without a reference, with scores ranging from 0 to 100.
Image quality: It provides a variety of quality inspection features, including but not limited to detection of screen glitch, black screen, jitter, noise, mosaic, low light, QR code, barcode, and mini program code.
Sound detection: It provides a variety of quality inspection features, including but not limited to detection of no voice, low voice, and high voice.

Description of Quality Inspection Template Parameters

Detection Type
Detection Item
Parameter Description
Format diagnosis
Format diagnosis
Media diagnosis is a cloud-based audio and video processing service for real-time diagnosis of media streams. It can be used to analyze exceptions in stream information, temporal information, stream status, container encapsulation, and decoding.
Quality score
No-reference scoring
When it is enabled, the system will activate the no-reference scoring and low-quality detection. Reference scores:
40-60: substandard quality;
60-80: average quality;
80-100: superior quality.
Two parameters can be configured: "threshold" and "detection interval". The threshold corresponds to the score given by the no-reference scoring feature, and a report will be generated when the score falls below the set threshold. The detection interval is measured in milliseconds (ms).
Image quality issues

Mosaic detection
Detects the presence of mosaic in the image. Two parameters can be configured: "threshold" and "detection interval". The threshold corresponds to the confidence level of mosaic detection, and a report will be generated when the detected confidence level is greater than or equal to the set threshold. The detection interval is measured in milliseconds (ms).
Screen glitch detection
Detects the presence of screen glitches in the image. Two parameters can be configured: "threshold" and "detection interval". The threshold corresponds to the confidence level of the screen glitch detection, and a report will be generated when the detected confidence level is greater than or equal to the set threshold. The detection interval is measured in milliseconds (ms).
Blur detection
Detects whether the image is blurry. Two parameters can be configured: "threshold" and "detection interval". The threshold serves as the evaluation benchmark during detection. The higher the threshold is set, the more likely an image is to be deemed blurry. The detection interval is measured in milliseconds (ms).
Jitter detection
Detects whether the image has jitter and ghosting issues. The "threshold" parameter can be configured. The threshold serves as the benchmark for determining jitter. The higher the threshold is, the more severe the jitter will be. A report will be generated when the jitter severity is greater than the set threshold.
Noise detection
Detects whether there is noise in the image. Two parameters can be configured: "threshold" and "detection interval". The threshold serves as the benchmark for determining noise. The higher the threshold is, the more obvious the noise will be. A report will be generated when the intensity level is greater than or equal to the set threshold. The detection interval is measured in milliseconds (ms).
QR code detection
Detects whether there is a QR code in the image. Two parameters can be configured: "threshold" and "detection interval". The threshold corresponds to the confidence level of QR code detection, and a report will be generated when the detected confidence level is greater than or equal to the set threshold. The detection interval is measured in milliseconds (ms).
Barcode detection
Detects whether there is a barcode in the image. Two parameters can be configured: "threshold" and "detection interval". The threshold corresponds to the confidence level of barcode detection, and a report will be generated when the detected confidence level is greater than or equal to the set threshold. The detection interval is measured in milliseconds (ms).
Mini program code detection
Detects whether there is a mini program code in the image. Two parameters can be configured: "threshold" and "detection interval". The threshold corresponds to the confidence level of mini program code detection, and a report will be generated when the detected confidence level is greater than or equal to the set threshold. The detection interval is measured in milliseconds (ms).
Solid color screen detection
Black and white edge detection
Detects whether there are black and white edges around the image.
The "detection interval" parameter can be configured in milliseconds (ms).
Solid color screen detection
Detects whether the screen is a solid color screen. Two parameters can be configured: "threshold" and "detection interval". The threshold is defined as the percentage of pixels of the same color, and a report will be generated when the detected percentage is greater than or equal to the set threshold. The detection interval is measured in milliseconds (ms).
Brightness detection
Low light detection
Detects whether the image is excessively dark. Two parameters can be configured: "threshold" and "detection interval". The threshold is defined as the percentage of low-light pixels, and a report will be generated when the detected percentage is greater than or equal to the set threshold. The detection interval is measured in milliseconds (ms).
Overexposure detection
Detects whether the image is overexposed. Two parameters can be configured: "threshold" and "detection interval". The threshold is defined as the percentage of overexposed pixels, and a report will be generated when the detected percentage is greater than or equal to the set threshold. The detection interval is measured in milliseconds (ms).
Sound detection
No voice detection
Detects whether the audio has no voice.
The "detection interval" parameter can be configured in milliseconds (ms).
Low voice detection
Detects whether the voice of the audio is too low. Two parameters can be configured: "threshold" and "detection interval". The threshold denotes the decibel level used to determine low voice, and a report will be generated when the audio decibel falls below the set value. The detection interval is measured in milliseconds (ms).
High voice detection
Detects whether the voice of the audio is too high. Two parameters can be configured: "threshold" and "detection interval".
The threshold denotes the decibel level used to determine high voice, and a report will be generated when the audio decibel exceeds the set value. The detection interval is measured in milliseconds (ms).