eggsywelsh / ephotoshop

EPhotoShop

Introduce

Based on android NDK development, the use of C language to dynamically change the image of the RGB value, to achieve the purpose of real-time picture processing. The initial built-in 9 in the transformation process, you can fork the project, and then run up to see the algorithm has been achieved.

How To Use

The main function please refer to EPhotoShop class, which encapsulates a variety of images to deal with the java method, the image processing algorithm using C language. Of course, you can refer to the EPhotoShop class, write your own implementation and algorithms, you can also be based on the expansion, which some see your needs.

call the image process method(by java)

Bitmap newBitMap = srcBitmap.copy(srcBitmap.getConfig(), true);
EPhotoShop.doGray(newBitMap);  // do the gray transformation
mIvImg.setImageBitmap(newBitMap);  // set the new BitMap into ImageView

Also,you can call the EPhotoShop class's method as you want.

Sample

We use the above method(EPhotoShop.doGray(newBitMap)) to look at the whole process

call method doGray

    public static Bitmap doGray(Bitmap inputBitmap) {
        Bitmap outputImage = inputBitmap;

        int width = inputBitmap.getWidth();
        int height = inputBitmap.getHeight();
        // get transformation area pixels
        int[] pixels = new int[width * height];
        outputImage.getPixels(pixels, 0, width, 0, 0, width, height);
        // do transformate
        pixels = NativeProcessor.applyGray(pixels, width, height);
        try {
            outputImage.setPixels(pixels, 0, width, 0, 0, width, height);
        } catch (Exception e) {
            e.printStackTrace();
        }
        return outputImage;
    }

call the process method NativeProcessor.applyGray (by C)

JNIEXPORT jintArray JNICALL
Java_com_eggsy_photoshop_NativeProcessor_applyGray(JNIEnv *env, jclass type,
                                                   jintArray pixels_, jint width,
                                                   jint height) {
    jint *pixels = env->GetIntArrayElements(pixels_, NULL);

    int newSize = width * height;

    int widthIndex = 0, heightIndex = 0, r, g, b, modif_color, curr_color;

    for (widthIndex = 0; widthIndex < width; widthIndex++) {
        for (heightIndex = 0; heightIndex < height; heightIndex++) {
            curr_color = pixels[heightIndex * width + widthIndex];
            r = red(curr_color);
            g = green(curr_color);
            b = blue(curr_color);
            modif_color = (int) (r * 0.3 + g * 0.59 + b * 0.11);
            pixels[heightIndex * width + widthIndex] = ARGB(alpha(curr_color), modif_color,
                                                            modif_color, modif_color);
        }
    }
    jintArray result = env->NewIntArray(newSize);
    env->SetIntArrayRegion(result, 0, newSize, pixels);

    env->ReleaseIntArrayElements(pixels_, pixels, 0);

    return result;
}

Through the above example, we know how to call the method from the java layer to deal with picture transformation, follow-up will have more algorithms to add to facilitate the use of everyone.