毕设&课设&项目&实训-图像缩放算法总汇，使用C++实现。包含创建的插值算法和基于变换的算法.zip

#include <jni.h> #include <string> #include <opencv2/opencv.hpp> #include <android/bitmap.h> #include <android/log.h> using namespace cv; #define LOG_TAG "AndroidScalingalgorithmforimage" #define LOGD(...) __android_log_print(ANDROID_LOG_DEBUG, LOG_TAG, __VA_ARGS__) jobject createBitmap(JNIEnv *env, int new_width, int new_height); extern "C" JNIEXPORT jstring JNICALL Java_com_kpa_image_MainActivity_stringFromJNI( JNIEnv *env, jobject /* this */) { std::string hello = "Hello from C++"; return env->NewStringUTF(hello.c_str()); } /** * 创建目标位图 * @param env * @param new_width 目标位图 宽 * @param new_height 高 * @return 目标位图 */ jobject createBitmap(JNIEnv *env, int new_width, int new_height) { jobject outputBitmap; jclass bitmapCls = env->FindClass("android/graphics/Bitmap"); jmethodID createBitmapMethod = env->GetStaticMethodID(bitmapCls, "createBitmap", "(IILandroid/graphics/Bitmap$Config;)Landroid/graphics/Bitmap;"); jstring configName = env->NewStringUTF("ARGB_8888"); jclass bitmapConfigClass = env->FindClass("android/graphics/Bitmap$Config"); jobject java_bitmap_config = env->CallStaticObjectMethod(bitmapConfigClass, env->GetStaticMethodID( bitmapConfigClass, "valueOf", "(Ljava/lang/String;)Landroid/graphics/Bitmap$Config;"), configName); outputBitmap = env->CallStaticObjectMethod(bitmapCls, createBitmapMethod, new_width, new_height, java_bitmap_config); return outputBitmap; } /** * 最近邻插值算法 */ extern "C" JNIEXPORT jobject JNICALL Java_com_kpa_image_MainActivity_scaleBitmapByNearestNeighbor(JNIEnv *env, jobject thiz, jobject inputBitmap, jfloat scale) { // bitmap 转换为openCV的Mat AndroidBitmapInfo bitmapInfo; Mat inputMat; Mat outputMat; jobject outputBitmap; if (scale == 1.0) { return inputBitmap; } // 将Android中的Bitmap 对象转为OpenCV 的Mat 对象 if (AndroidBitmap_getInfo(env, inputBitmap, &bitmapInfo) < 0) { LOGD("转换失败"); return NULL; } LOGD("转换成功"); LOGD("图片的格式 %d", bitmapInfo.format); void *inputPixels; //获取bitmap 的像素数据（锁定像素缓冲区并检索指向它的指针。这样做可确保像素在应用调用 AndroidBitmap_unlockPixels 之前不会移动） if (AndroidBitmap_lockPixels(env, inputBitmap, &inputPixels) < 0) { LOGD("获取像素数据失败"); return NULL; } // 创建OpenCV的Mat对象 LOGD("height = %d, width = %d", bitmapInfo.height, bitmapInfo.width); inputMat.create(bitmapInfo.height, bitmapInfo.width, CV_8UC4); // 将Bitmap的像素数据拷贝到Mat对象中 memcpy(inputMat.data, inputPixels, inputMat.rows * inputMat.step); //解锁Bitmap的像素数据 AndroidBitmap_unlockPixels(env, inputBitmap); // 将图像缩小到原来的一半大小 // int new_width = inputMat.cols / 2; // int new_height = inputMat.rows / 2; // 根据缩放比例计算输出图像的宽高 int new_width = static_cast<int>(inputMat.cols * scale); int new_height = static_cast<int>(inputMat.cols * scale); //创建输出的Mat对象 outputMat.create(new_height, new_width, CV_8UC4); // 遍历每个像素 for (int i = 0; i < new_height; i++) { for (int j = 0; j < new_width; j++) { //获取原始像素位置 int orig_i = static_cast<int>(i / scale); int orig_j = static_cast<int>(j / scale); //将原始像素拷贝到输出Mat对象中 outputMat.at<Vec4b>(i, j) = inputMat.at<Vec4b>(orig_i, orig_j); } } LOGD("图像缩放完成"); outputBitmap = createBitmap(env, new_width, new_height); LOGD("将OpenCV的Mat对象转换成Java中的Bitmap对象完成"); void *outputPixels; int result = AndroidBitmap_lockPixels(env, outputBitmap, &outputPixels); if (result < 0) { LOGD("outputPixels执行失败 %d", result); return NULL; } LOGD("outputPixels执行完成"); memcpy(outputPixels, outputMat.data, outputMat.rows * outputMat.step); AndroidBitmap_unlockPixels(env, outputBitmap); LOGD("执行完成"); // 返回Java中的Bitmap对象 return outputBitmap; } /** * 双线性插值 */ extern "C" JNIEXPORT jobject JNICALL Java_com_kpa_image_MainActivity_scaleBitmapByBilinear(JNIEnv *env, jobject thiz, jobject input_bitmap, jfloat scale) { if (scale == 1.0) { return input_bitmap; } // 获取源位图和目标位图的信息 AndroidBitmapInfo bitmapInfo; void *pixels = nullptr; //将Android中的Bitmap 对象转为OpenCV 的Mat 对象 if (AndroidBitmap_getInfo(env, input_bitmap, &bitmapInfo) < 0) { return NULL; } //获取原位图的像素 if (AndroidBitmap_lockPixels(env, input_bitmap, &pixels) < 0) { return NULL; } //解锁Bitmap的像素数据 AndroidBitmap_unlockPixels(env, input_bitmap); LOGD("height = %d, width = %d", bitmapInfo.height, bitmapInfo.width); //按比例计算目标图像的宽高 int newWidth = static_cast<int>(bitmapInfo.width * scale); int newHeight = static_cast<int>(bitmapInfo.height * scale); LOGD("newHeight = %d, newWidth = %d", newHeight, newWidth); // 原位图像素数组 uint32_t *srcPixels = static_cast<uint32_t *>(pixels); // 目标位图像素数组 uint32_t *dstPixels = new uint32_t[newWidth * newHeight]; //位图的宽度和新位图的宽度都是从1开始计数的，而像素索引是从0开始计数的。因此，要计算像素索引之间的比率，需要将宽度减一。 float xRatio = static_cast<float>(bitmapInfo.width - 1) / static_cast<float>(newWidth - 1); float yRatio = static_cast<float>(bitmapInfo.height - 1) / static_cast<float>(newHeight - 1); for (int y = 0; y < newHeight; ++y) { for (int x = 0; x < newWidth; ++x) { float gx = x * xRatio; float gy = y * yRatio; int gxi = static_cast<int>(gx); int gyi = static_cast<int>(gy); float fracx = gx - gxi; float fracy = gy - gyi; uint32_t c00 = srcPixels[gyi * bitmapInfo.width + gxi]; uint32_t c10 = srcPixels[gyi * bitmapInfo.width + gxi + 1]; uint32_t c01 = srcPixels[(gyi + 1) * bitmapInfo.width + gxi]; uint32_t c11 = srcPixels[(gyi + 1) * bitmapInfo.width + gxi + 1]; int r = static_cast<int>((1 - fracx) * (1 - fracy) * (c00 >> 16 & 0xff) + fracx * (1 - fracy) * (c10 >> 16 & 0xff) + (1 - fracx) * fracy * (c01 >> 16 & 0xff) + fracx * fracy * (c11 >> 16 & 0xff)); int g = static_cast<int>((1 - fracx) * (1 - fracy) * (c00 >> 8 & 0xff) + fracx * (1 - fracy) * (c10 >> 8 & 0xff) + (1 - fracx) * fracy * (c01 >> 8 & 0xff) + fracx * fracy * (c11 >> 8 & 0xff)); int b = static_cast<int>((1 - fracx) * (1 - fracy) * (c00 & 0xff) + fracx * (1 - fracy) * (c10 & 0xff) + (1 - f