How to Boost Mobile Video Quality with Brightness and Color Enhancement Techniques

1. Introduction

As player architectures evolve, video post‑processing becomes a key factor for improving user experience. Modern players increasingly add post‑processing chains to enhance image quality and create immersive visual effects.

2. Video Enhancement (Brightness and Color)

2.1 What is video enhancement technology

Video enhancement refers to a series of techniques that improve visual quality without altering the original content, applicable to playback, editing, transmission, and storage.

2.2 Common video enhancement techniques

Mobile practice focuses on brightness and color enhancement for Android/iOS playback scenarios.

2.3 Brightness enhancement

Brightness enhancement effect diagram (left: original, right: enhanced).

2.3.1 Technical selection

Linear brightness enhancement (gain):

color.rgb = color.rgb * gain; // multiplication enhancement
color.rgb = color.rgb + offset; // addition enhancement

Converting RGB to YUV and adjusting the Y channel can avoid color bias.

Y = 0.299*R + 0.587*G + 0.114*B;
Y_new = Y * gain;

Histogram equalization distributes pixel luminance across the full range, improving contrast for low‑contrast images.

Gamma transformation (power law): color.rgb = pow(color.rgb, vec3(gamma)); Gamma < 1 brightens the image (lifts shadows); gamma > 1 darkens it (compresses highlights). This method is computationally cheap and suitable for real‑time playback.

2.3.2 Underlying principle

Gamma‑based adjustment lifts darker pixels more than brighter ones, preserving highlight details while increasing overall exposure.

Encoding uses gamma, so post‑processing must follow gamma‑space rules to maintain natural perception.

2.4 Color enhancement

Color enhancement effect diagram (left: original, right: enhanced).

2.4.1 Goals

Enhance color perception – increase vividness and visual appeal.

Highlight the subject – boost color contrast between foreground and background.

Restore realistic colors – correct color loss in skin, foliage, sky, etc.

2.4.2 Technical selection

Two research directions: traditional SDR color enhancement and SDR‑to‑HDR simulation. Implementation approaches: non‑neural algorithms (including LUTs) and neural‑network models. The neural approach was omitted due to high computational cost on mobile devices.

2.4.2.1 Color three elements

Hue, Saturation, Lightness (or Value) together define a color’s visual characteristics.

In color enhancement, saturation is the primary target; brightness may be adjusted as a secondary aid, while hue is usually left unchanged.

2.4.2.2 Choosing a color space

RGB

HSV

LCH/LAB

2.4.2.3 Based on RGB space

Direct RGB adjustment using luma interpolation:

luma = 0.2126*r + 0.7152*g + 0.0722*b;
color.rgb = mix(vec3(luma), color.rgb, k);

This can cause over‑saturation and loss of detail in already vivid regions.

2.4.2.4 Based on HSV space

Adjusting saturation in HSV modifies color vividness without affecting hue or value.

Potential issues include over‑saturation in already vivid areas and color banding in low‑saturation regions.

2.4.2.5 Skin‑tone protection

After HSV‑based enhancement, skin tones can become distorted; a protection step based on HSV skin‑tone detection and Gaussian probability weighting mitigates this.

Convert RGB to HSV.

Compute skin‑tone probability using hue distance and Gaussian weighting.

Apply a mask (e.g., probability > 0.95) and attenuate enhancement strength in skin regions.

2.4.3 Effect comparison

HSV‑based adjustment yields more natural colors.

RGB‑based adjustment is more vivid but can introduce color bias.

Overall, HSV is preferred for broader source compatibility.

3. Summary and Outlook

The study focuses on mobile video enhancement, validating brightness and color enhancement algorithms that have been deployed in the "Haokan" app. Future work includes scene‑specific optimization, lightweight model acceleration, and further real‑time performance improvements.