我的C++代码运行在Win 10上，自构建OpenCV 4.5.3.，检查WITH_OPENCL标志。
通过OpenCL的“透明API”，使用UMat代替Mat不会带来任何性能改进。
从我所读到的https://jeanvitor.com/opencv-opencl-umat-performance/，我期望至少有一个轻微的性能改善时，使用UMat，而不是Mat，即使在我的笔记本电脑英特尔高清图形520。
对UMat和Mat执行resize、cvtColor和bilateralFilter不会显示任何差异。
ocl::haveOpenCL和ocl::useOpenCL都返回真。
唯一Device::TYPE_GPU设备的查询值为：
name: Intel(R) HD Graphics 520
extensions: cl_khr_3d_image_writes cl_khr_byte_addressable_store cl_khr_fp16 cl_khr_depth_images cl_khr_global_int32_base_atomics cl_khr_global_int32_extended_atomics cl_khr_icd cl_khr_image2d_from_buffer cl_khr_local_int32_base_atomics cl_khr_local_int32_extended_atomics cl_intel_subgroups cl_intel_required_subgroup_size cl_intel_subgroups_short cl_khr_spir cl_intel_accelerator cl_intel_media_block_io cl_intel_driver_diagnostics cl_intel_device_side_avc_motion_estimation cl_khr_priority_hints cl_khr_throttle_hints cl_khr_create_command_queue cl_khr_fp64 cl_khr_subgroups cl_khr_il_program cl_khr_mipmap_image cl_khr_mipmap_image_writes cl_intel_planar_yuv cl_intel_packed_yuv cl_intel_motion_estimation cl_intel_advanced_motion_estimation cl_khr_gl_sharing cl_khr_gl_depth_images cl_khr_gl_event cl_khr_gl_msaa_sharing cl_intel_dx9_media_sharing cl_khr_dx9_media_sharing cl_khr_d3d10_sharing cl_khr_d3d11_sharing cl_intel_d3d11_nv12_media_sharing cl_intel_simultaneous_sharing个
version: OpenCL 2.1 NEO
OpenCLVersion: OpenCL 2.1 NEO
OpenCL_C_Version: OpenCL C 2.0
我也在其他机器上试过，性能没有差异。
我错过了什么，有什么想法或提示吗？