- 操作系统:ubuntu22.04
- OpenCV版本:OpenCV4.9
- IDE:Visual Studio Code
- 编程语言:C++11
算法描述
该函数用于计算四维浮点向量(float4类型)的双曲余弦值,作用于CUDA设备端。双曲余弦函数定义为cosh(x) = (eˣ + e⁻ˣ)/2,具有偶函数性质(cosh(-x) = cosh(x))
函数原型
__device__ __forceinline__ float4 cv::cudev::cosh ( const float4 & a )
参数
参数 | 类型 | 描述 |
---|---|---|
a | const float4& | 输入向量,每个分量独立计算双曲余弦 |
返回值
返回float4类型向量,其每个分量为输入向量对应分量的双曲余弦值,值域为[1, +∞)
应用场景
适用于GPU加速的数学计算、图像处理(如非线性滤波)和物理模拟等领域。
代码示例
#include <opencv2/opencv.hpp>
#include <opencv2/cudev/common.hpp>
#include <opencv2/cudev/util/vec_math.hpp>__global__ void kernel_cosh(const float4* input, float4* output, int size) {int idx = blockIdx.x * blockDim.x + threadIdx.x;if (idx < size) {output[idx] = cv::cudev::cosh(input[idx]);}
}int main() {const int N = 4;float4 h_input[N] = {{0.0f, 1.0f, 2.0f, 3.0f},{-1.0f, -2.0f, -3.0f, -4.0f},{0.5f, 1.5f, 2.5f, 3.5f},{-0.5f, -1.5f, -2.5f, -3.5f}};float4 h_output[N];// 分配设备内存float4* d_input;float4* d_output;cudaMalloc(&d_input, N * sizeof(float4));cudaMalloc(&d_output, N * sizeof(float4));// 拷贝数据到设备cudaMemcpy(d_input, h_input, N * sizeof(float4), cudaMemcpyHostToDevice);// 调用核函数dim3 block(4);dim3 grid(1);kernel_cosh<<<grid, block>>>(d_input, d_output, N);// 拷贝结果回主机cudaMemcpy(h_output, d_output, N * sizeof(float4), cudaMemcpyDeviceToHost);// 打印结果for (int i = 0; i < N; ++i) {printf("cosh(%.1f, %.1f, %.1f, %.1f) = (%.6f, %.6f, %.6f, %.6f)\n",h_input[i].x, h_input[i].y, h_input[i].z, h_input[i].w,h_output[i].x, h_output[i].y, h_output[i].z, h_output[i].w);}// 释放内存cudaFree(d_input);cudaFree(d_output);return 0;
}
运行结果
cosh(0.0, 1.0, 2.0, 3.0) = (1.000000, 1.543081, 3.762196, 10.067662)
cosh(-1.0, -2.0, -3.0, -4.0) = (1.543081, 3.762196, 10.067662, 27.308231)
cosh(0.5, 1.5, 2.5, 3.5) = (1.127626, 2.352410, 6.132289, 16.572824)
cosh(-0.5, -1.5, -2.5, -3.5) = (1.127626, 2.352410, 6.132289, 16.572824)