Merge pull request #28442 from Anemptyship:optimize/resize-parallel

optimize(dnn): parallelize Resize layer implementation #28442

### optimize(dnn): parallelize Resize layer implementation

This PR addresses the `TODO` in `modules/dnn/src/layers/resize_layer.cpp` regarding the slow implementation of the `Resize` layer when using `opencv_linear` or `nearest` interpolation with specific configurations.

### Changes
- Replaced the serial nested loop (batch x channels) with `cv::parallel_for_`.
- This allows OpenCV to utilize multi-threading for the resizing operation, which was previously single-threaded for these specific paths.

### Performance Results
Benchmark run on 20-thread CPU (AVX2):

| Test Case | Resolution Change | Before (ms) | After (ms) | Speedup |
| :--- | :--- | :--- | :--- | :--- |
| **Upsample Linear** | 64x64 -> 128x128 | 1.47 ms | **0.18 ms** | **~8.1x** |
| **Downsample Nearest** | 128x128 -> 64x64 | 1.55 ms | **0.45 ms** | **~3.4x** |

**Test Configuration:**
- **Upsample**: `[4, 64, 64, 64]` -> `[4, 64, 128, 128]` (Factor: 2.0, Linear)
- **Downsample**: `[4, 128, 128, 128]` -> `[4, 128, 64, 64]` (Factor: 0.5, Nearest)

### Pull Request Readiness Checklist

See details at https://github.com/opencv/opencv/wiki/How_to_contribute#making-a-good-pull-request

- [x] I agree to contribute to the project under Apache 2 License.
- [x] To the best of my knowledge, the proposed patch is not based on a code under GPL or another license that is incompatible with OpenCV
- [x] The PR is proposed to the proper branch
- [x] There is a reference to the original bug report and related work
- [ ] There is accuracy test, performance test and test data in opencv_extra repository, if applicable
      - Added `modules/dnn/perf/perf_resize.cpp` to verify performance gains.
- [ ] The feature is well documented and sample code can be built with the project CMake

modules/dnn/perf/perf_resize.cpp

@@ -0,0 +1,65 @@
+// This file is part of OpenCV project.
+// It is subject to the license terms in the LICENSE file found in the top-level directory
+// of this distribution and at http://opencv.org/license.html.
+
+#include "perf_precomp.hpp"
+
+namespace opencv_test {
+
+struct Layer_Resize : public TestBaseWithParam<tuple<Backend, Target>>
+{
+    void test_layer(const std::vector<int>& inpShape, int outH, int outW, const String& interp)
+    {
+        int backendId = get<0>(GetParam());
+        int targetId = get<1>(GetParam());
+
+        Mat input(inpShape, CV_32FC1);
+        randu(input, 0.f, 1.f);
+
+        Net net;
+        LayerParams lp;
+        lp.type = "Resize";
+        lp.name = "testLayer";
+        lp.set("interpolation", interp);
+        lp.set("width", outW);
+        lp.set("height", outH);
+
+
+
+        int id = net.addLayerToPrev(lp.name, lp.type, lp);
+        net.connect(0, 0, id, 0);
+
+        // warmup
+        {
+            net.setInputsNames({"data"});
+            net.setInput(input, "data");
+            net.setPreferableBackend(backendId);
+            net.setPreferableTarget(targetId);
+            Mat out = net.forward();
+        }
+
+        TEST_CYCLE()
+        {
+            Mat res = net.forward();
+        }
+
+        SANITY_CHECK_NOTHING();
+    }
+};
+
+PERF_TEST_P_(Layer_Resize, Resize_Upsample_Linear)
+{
+    // N=4, C=64, H=64, W=64 -> 128x128 (x2 upsample)
+    // Common in segmentation/detection heads
+    test_layer({4, 64, 64, 64}, 128, 128, "opencv_linear");
+}
+
+PERF_TEST_P_(Layer_Resize, Resize_Downsample_Nearest)
+{
+    // N=4, C=128, H=128, W=128 -> 64x64 (x0.5 downsample)
+    test_layer({4, 128, 128, 128}, 64, 64, "nearest");
+}
+
+INSTANTIATE_TEST_CASE_P(/**/, Layer_Resize, dnnBackendsAndTargets());
+
+} // namespace opencv_test

modules/dnn/src/layers/resize_layer.cpp

@@ -244,15 +244,20 @@ public:
         {
             // INTER_LINEAR Resize mode does not support INT8 inputs
             InterpolationFlags mode = interpolation == "nearest" ? INTER_NEAREST : INTER_LINEAR;
-            // [TODO] this is a really slow approach; need to rewrite it completely.
-            for (size_t n = 0; n < inputs[0].size[0]; ++n)
-            {
-                for (size_t ch = 0; ch < inputs[0].size[1]; ++ch)
+
+            size_t nbatch = inputs[0].size[0];
+            size_t nch = inputs[0].size[1];
+            size_t total_planes = nbatch * nch;
+
+            parallel_for_(Range(0, (int)total_planes), [&](const Range& range){
+                for (int i = range.start; i < range.end; ++i)
                 {
+                    int n = i / nch;
+                    int ch = i % nch;
                     resize(getPlane(inp, n, ch), getPlane(out, n, ch),
                            Size(outWidth, outHeight), 0, 0, mode);
                 }
-            }
+            });
         }
         else if (interpolation == "nearest")
         {