init project

zch-90 · Aug 30, 2020 · c71b8ca · c71b8ca · tuns11 · May 14, 2023
1 parent 2ca85dc
commit c71b8ca
Show file tree

Hide file tree

Showing 3 changed files with 389 additions and 4 deletions.
diff --git a/Readme.md b/Readme.md
@@ -1,4 +1,9 @@
-# 1. How to use (GLSL / MPV) 
-## 1.1 install MPV
-## 1.2 Download the .glsl shader files here
-## 1.3 Copy the .glsl files to %AppData%\mpv\shaders for Windows
+# AI-video-enhance
+This repository is a set of open-source, high-quality real-time video enhance algorithms that can be implemented in any programming language.
+
+# How to use (GLSL / MPV) 
+1. install MPV
+2. Download the .glsl shader files here
+3. Copy the .glsl files to %AppData%\mpv\shaders for Windows
+
+
diff --git a/result/1.mp4 b/result/1.mp4
@@ -0,0 +1,190 @@
+//Anime4K v3.1 GLSL
+
+// This is free and unencumbered software released into the public domain.
+
+// Anyone is free to copy, modify, publish, use, compile, sell, or
+// distribute this software, either in source code form or as a compiled
+// binary, for any purpose, commercial or non-commercial, and by any
+// means.
+
+// In jurisdictions that recognize copyright laws, the author or authors
+// of this software dedicate any and all copyright interest in the
+// software to the public domain. We make this dedication for the benefit
+// of the public at large and to the detriment of our heirs and
+// successors. We intend this dedication to be an overt act of
+// relinquishment in perpetuity of all present and future rights to this
+// software under copyright law.
+
+// THE SOFTWARE IS PROVIDED "AS IS", WITHOUT WARRANTY OF ANY KIND,
+// EXPRESS OR IMPLIED, INCLUDING BUT NOT LIMITED TO THE WARRANTIES OF
+// MERCHANTABILITY, FITNESS FOR A PARTICULAR PURPOSE AND NONINFRINGEMENT.
+// IN NO EVENT SHALL THE AUTHORS BE LIABLE FOR ANY CLAIM, DAMAGES OR
+// OTHER LIABILITY, WHETHER IN AN ACTION OF CONTRACT, TORT OR OTHERWISE,
+// ARISING FROM, OUT OF OR IN CONNECTION WITH THE SOFTWARE OR THE USE OR
+// OTHER DEALINGS IN THE SOFTWARE.
+
+// For more information, please refer to <https://unlicense.org>
+
+//!DESC Anime4K-v3.1-DarkLines-Kernel(X)
+//!HOOK NATIVE
+//!BIND HOOKED
+
+vec3 rgb2hsv(vec3 c)
+{
+    vec4 K = vec4(0.0, -1.0 / 3.0, 2.0 / 3.0, -1.0);
+    vec4 p = mix(vec4(c.bg, K.wz), vec4(c.gb, K.xy), step(c.b, c.g));
+    vec4 q = mix(vec4(p.xyw, c.r), vec4(c.r, p.yzx), step(p.x, c.r));
+
+    float d = q.x - min(q.w, q.y);
+    float e = 1.0e-10;
+    return vec3(abs(q.z + (q.w - q.y) / (6.0 * d + e)), d / (q.x + e), q.x);
+}
+
+vec3 hsv2rgb(vec3 c)
+{
+    vec4 K = vec4(1.0, 2.0 / 3.0, 1.0 / 3.0, 3.0);
+    vec3 p = abs(fract(c.xxx + K.xyz) * 6.0 - K.www);
+    return c.z * mix(K.xxx, clamp(p - K.xxx, 0.0, 1.0), c.y);
+}
+
+vec3 sat(vec3 img)
+{
+    img = img / 255.0;
+    float Increment  = 0.03;
+    float img_min = min(img.r, min(img.g, img.b));
+    float img_max = max(img.r, max(img.g, img.b));
+
+    float Delta = (img_max - img_min);
+    float value = (img_max + img_min);
+
+    float L = value/2.0;
+
+    float mask_1 = 0.0;
+    if(L < 0.5) {
+        mask_1 = 1.0;
+    } else {
+        mask_1 = 0.0;
+    }
+
+    float s1 = Delta / (value + 0.001);
+    float s2 = Delta / (2.0 - value + 0.001);
+    float s = s1*mask_1 + s2*(1.0 - mask_1);
+
+
+    if(Increment >= 0.0) {
+        float temp = Increment + s;
+        float mask_2 = 0.0;
+        if(temp > 1.0){
+            mask_2 = 1.0;
+        } else {
+            mask_2 = 0.0;
+        }
+        float alpha_1 = s;
+        float alpha_2 = s*0.0 + 1.0 - Increment;
+        float alpha = alpha_1*mask_2 + alpha_2*(1.0 - mask_2);
+        alpha = 1.0/(alpha + 0.001) - 1.0;
+        img.r = img.r + (img.r - L) * alpha;
+        img.g = img.g + (img.g - L) * alpha;
+        img.b = img.b + (img.b - L) * alpha;
+    } else {
+        float alpha = Increment;
+        img.r = img.r + (img.r - L) * alpha;
+        img.g = img.g + (img.g - L) * alpha;
+        img.b = img.b + (img.b - L) * alpha;        
+    }
+
+
+    vec3 img_out = img;
+    float mask_r_1 = 0.0;
+    float mask_r_2 = 0.0;
+    float mask_g_1 = 0.0;
+    float mask_g_2 = 0.0;
+    float mask_b_1 = 0.0;
+    float mask_b_2 = 0.0;
+
+    if(img_out.r < 0.0){
+        mask_r_1 = 1.0;
+    } else {
+        mask_r_1 = 0.0;
+    }
+    if(img_out.r > 1.0){
+        mask_r_2 = 1.0;
+    } else {
+        mask_r_2 = 0.0;
+    }
+
+    if(img_out.g < 0.0){
+        mask_g_1 = 1.0;
+    } else {
+        mask_g_1 = 0.0;
+    }
+    if(img_out.g > 1.0){
+        mask_g_2 = 1.0;
+    } else {
+        mask_g_2 = 0.0;
+    }
+
+
+    if(img_out.b < 0.0){
+        mask_b_1 = 1.0;
+    } else {
+        mask_b_1 = 0.0;
+    }
+    if(img_out.b > 1.0){
+        mask_b_2 = 1.0;
+    } else {
+        mask_b_2 = 0.0;
+    }
+
+
+    //img_out = img_out * (1.0-mask_r_1, 1.0-mask_g_1, 1.0-mask_b_1);
+    img_out.r = img_out.r*(1.0-mask_r_1);
+    img_out.g = img_out.g*(1.0-mask_g_1);
+    img_out.b = img_out.b*(1.0-mask_b_1);
+    //img_out = img_out * (1.0-mask_r_2, 1.0-mask_g_2, 1.0-mask_b_2) + (mask_r_2, mask_g_2, mask_b_2);
+    img_out.r = img_out.r*(1.0-mask_r_2);
+    img_out.g = img_out.g*(1.0-mask_g_2);
+    img_out.b = img_out.b*(1.0-mask_b_2);
+    img_out += (mask_r_2, mask_g_2, mask_b_2);
+    img_out = img_out * 255.0;
+
+    return img_out;
+}
+
+#define L_tex HOOKED_tex
+
+#define SIGMA 1
+
+float gaussian(float x, float s, float m) {
+	return (1 / (s * sqrt(2 * 3.14159))) * exp(-0.5 * pow(abs(x - m) / s, 2.0));
+}
+
+float lumGaussian(vec2 pos, vec2 d) {
+	float s = SIGMA * HOOKED_size.y / 1080;
+	float kernel_size = s * 2 + 1;
+
+	float g = (L_tex(pos).x) * gaussian(0, s, 0);
+	float gn = gaussian(0, s, 0);
+
+	g += (L_tex(pos - d).x + L_tex(pos + d).x) * gaussian(1, s, 0);
+	gn += gaussian(1, s, 0) * 2;
+
+	for (int i=2; i<kernel_size; i++) {
+		g += (L_tex(pos - (d * i)).x + L_tex(pos + (d * i)).x) * gaussian(i, s, 0);
+		gn += gaussian(i, s, 0) * 2;
+	}
+
+	return g / gn;
+}
+
+float getLum(vec4 rgb) {
+	return (rgb.r + rgb.r + rgb.g + rgb.g + rgb.g + rgb.b) / 6.0;
+}
+
+vec4 hook() { //Save lum on OUTPUT
+	vec4 rgb = HOOKED_tex(HOOKED_pos);
+    rgb.rgb = sat(rgb.rgb);
+    //rgb.rgb = rgb.rgb / 255.0;
+    //rgb.rgb = rgb.rgb * 255.0;
+    return rgb; //vec4(255.0); //vec4(lum);
+}
diff --git a/video/1.mp4 b/video/1.mp4
@@ -0,0 +1,190 @@
+//Anime4K v3.1 GLSL
+
+// This is free and unencumbered software released into the public domain.
+
+// Anyone is free to copy, modify, publish, use, compile, sell, or
+// distribute this software, either in source code form or as a compiled
+// binary, for any purpose, commercial or non-commercial, and by any
+// means.
+
+// In jurisdictions that recognize copyright laws, the author or authors
+// of this software dedicate any and all copyright interest in the
+// software to the public domain. We make this dedication for the benefit
+// of the public at large and to the detriment of our heirs and
+// successors. We intend this dedication to be an overt act of
+// relinquishment in perpetuity of all present and future rights to this
+// software under copyright law.
+
+// THE SOFTWARE IS PROVIDED "AS IS", WITHOUT WARRANTY OF ANY KIND,
+// EXPRESS OR IMPLIED, INCLUDING BUT NOT LIMITED TO THE WARRANTIES OF
+// MERCHANTABILITY, FITNESS FOR A PARTICULAR PURPOSE AND NONINFRINGEMENT.
+// IN NO EVENT SHALL THE AUTHORS BE LIABLE FOR ANY CLAIM, DAMAGES OR
+// OTHER LIABILITY, WHETHER IN AN ACTION OF CONTRACT, TORT OR OTHERWISE,
+// ARISING FROM, OUT OF OR IN CONNECTION WITH THE SOFTWARE OR THE USE OR
+// OTHER DEALINGS IN THE SOFTWARE.
+
+// For more information, please refer to <https://unlicense.org>
+
+//!DESC Anime4K-v3.1-DarkLines-Kernel(X)
+//!HOOK NATIVE
+//!BIND HOOKED
+
+vec3 rgb2hsv(vec3 c)
+{
+    vec4 K = vec4(0.0, -1.0 / 3.0, 2.0 / 3.0, -1.0);
+    vec4 p = mix(vec4(c.bg, K.wz), vec4(c.gb, K.xy), step(c.b, c.g));
+    vec4 q = mix(vec4(p.xyw, c.r), vec4(c.r, p.yzx), step(p.x, c.r));
+
+    float d = q.x - min(q.w, q.y);
+    float e = 1.0e-10;
+    return vec3(abs(q.z + (q.w - q.y) / (6.0 * d + e)), d / (q.x + e), q.x);
+}
+
+vec3 hsv2rgb(vec3 c)
+{
+    vec4 K = vec4(1.0, 2.0 / 3.0, 1.0 / 3.0, 3.0);
+    vec3 p = abs(fract(c.xxx + K.xyz) * 6.0 - K.www);
+    return c.z * mix(K.xxx, clamp(p - K.xxx, 0.0, 1.0), c.y);
+}
+
+vec3 sat(vec3 img)
+{
+    img = img / 255.0;
+    float Increment  = 0.03;
+    float img_min = min(img.r, min(img.g, img.b));
+    float img_max = max(img.r, max(img.g, img.b));
+
+    float Delta = (img_max - img_min);
+    float value = (img_max + img_min);
+
+    float L = value/2.0;
+
+    float mask_1 = 0.0;
+    if(L < 0.5) {
+        mask_1 = 1.0;
+    } else {
+        mask_1 = 0.0;
+    }
+
+    float s1 = Delta / (value + 0.001);
+    float s2 = Delta / (2.0 - value + 0.001);
+    float s = s1*mask_1 + s2*(1.0 - mask_1);
+
+
+    if(Increment >= 0.0) {
+        float temp = Increment + s;
+        float mask_2 = 0.0;
+        if(temp > 1.0){
+            mask_2 = 1.0;
+        } else {
+            mask_2 = 0.0;
+        }
+        float alpha_1 = s;
+        float alpha_2 = s*0.0 + 1.0 - Increment;
+        float alpha = alpha_1*mask_2 + alpha_2*(1.0 - mask_2);
+        alpha = 1.0/(alpha + 0.001) - 1.0;
+        img.r = img.r + (img.r - L) * alpha;
+        img.g = img.g + (img.g - L) * alpha;
+        img.b = img.b + (img.b - L) * alpha;
+    } else {
+        float alpha = Increment;
+        img.r = img.r + (img.r - L) * alpha;
+        img.g = img.g + (img.g - L) * alpha;
+        img.b = img.b + (img.b - L) * alpha;        
+    }
+
+
+    vec3 img_out = img;
+    float mask_r_1 = 0.0;
+    float mask_r_2 = 0.0;
+    float mask_g_1 = 0.0;
+    float mask_g_2 = 0.0;
+    float mask_b_1 = 0.0;
+    float mask_b_2 = 0.0;
+
+    if(img_out.r < 0.0){
+        mask_r_1 = 1.0;
+    } else {
+        mask_r_1 = 0.0;
+    }
+    if(img_out.r > 1.0){
+        mask_r_2 = 1.0;
+    } else {
+        mask_r_2 = 0.0;
+    }
+
+    if(img_out.g < 0.0){
+        mask_g_1 = 1.0;
+    } else {
+        mask_g_1 = 0.0;
+    }
+    if(img_out.g > 1.0){
+        mask_g_2 = 1.0;
+    } else {
+        mask_g_2 = 0.0;
+    }
+
+
+    if(img_out.b < 0.0){
+        mask_b_1 = 1.0;
+    } else {
+        mask_b_1 = 0.0;
+    }
+    if(img_out.b > 1.0){
+        mask_b_2 = 1.0;
+    } else {
+        mask_b_2 = 0.0;
+    }
+
+
+    //img_out = img_out * (1.0-mask_r_1, 1.0-mask_g_1, 1.0-mask_b_1);
+    img_out.r = img_out.r*(1.0-mask_r_1);
+    img_out.g = img_out.g*(1.0-mask_g_1);
+    img_out.b = img_out.b*(1.0-mask_b_1);
+    //img_out = img_out * (1.0-mask_r_2, 1.0-mask_g_2, 1.0-mask_b_2) + (mask_r_2, mask_g_2, mask_b_2);
+    img_out.r = img_out.r*(1.0-mask_r_2);
+    img_out.g = img_out.g*(1.0-mask_g_2);
+    img_out.b = img_out.b*(1.0-mask_b_2);
+    img_out += (mask_r_2, mask_g_2, mask_b_2);
+    img_out = img_out * 255.0;
+
+    return img_out;
+}
+
+#define L_tex HOOKED_tex
+
+#define SIGMA 1
+
+float gaussian(float x, float s, float m) {
+	return (1 / (s * sqrt(2 * 3.14159))) * exp(-0.5 * pow(abs(x - m) / s, 2.0));
+}
+
+float lumGaussian(vec2 pos, vec2 d) {
+	float s = SIGMA * HOOKED_size.y / 1080;
+	float kernel_size = s * 2 + 1;
+
+	float g = (L_tex(pos).x) * gaussian(0, s, 0);
+	float gn = gaussian(0, s, 0);
+
+	g += (L_tex(pos - d).x + L_tex(pos + d).x) * gaussian(1, s, 0);
+	gn += gaussian(1, s, 0) * 2;
+
+	for (int i=2; i<kernel_size; i++) {
+		g += (L_tex(pos - (d * i)).x + L_tex(pos + (d * i)).x) * gaussian(i, s, 0);
+		gn += gaussian(i, s, 0) * 2;
+	}
+
+	return g / gn;
+}
+
+float getLum(vec4 rgb) {
+	return (rgb.r + rgb.r + rgb.g + rgb.g + rgb.g + rgb.b) / 6.0;
+}
+
+vec4 hook() { //Save lum on OUTPUT
+	vec4 rgb = HOOKED_tex(HOOKED_pos);
+    rgb.rgb = sat(rgb.rgb);
+    //rgb.rgb = rgb.rgb / 255.0;
+    //rgb.rgb = rgb.rgb * 255.0;
+    return rgb; //vec4(255.0); //vec4(lum);
+}