HED: Add Holistically-nested Edge Detection module

[sasankgit]

Description

This PR introduces the Holistically-nested Edge Detection (HED) algorithm as a new module (opencv_hed) in opencv_contrib.

HED (Xie & Tu, 2015) is a deep-learning-based edge detection system that utilizes a pretrained VGG network to perform holistically-nested image-to-image edge learning. It learns rich hierarchical features to detect semantically meaningful boundaries rather than raw pixel contrast changes.

Key Features:

• Implements the cv::hed::HEDDetector class which loads the Caffe-based model (hed_pretrained_bsds.caffemodel and deploy.prototxt) and performs the forward pass.
• Returns a 32-bit single-channel floating-point edge map with values normalized in the [0, 1] range.
• Python wrappers are enabled via the module build configuration.
• Unit tests are added to verify the interface and output shape. Note that tests use cvtest::findDataFile to locate the model files (deploy.prototxt and hed_pretrained_bsds.caffemodel), which are submitted in a separate PR to the main opencv repository under data/.

References:

• Paper: Xie, S., & Tu, Z. (2015). Holistically-nested edge detection. In Proceedings of the IEEE international conference on computer vision (pp. 1395-1403). arXiv:1504.06375.

Pull Request Readiness Checklist

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

• I agree to contribute to the project under Apache 2 License.
• 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
• The PR is proposed to the proper branch
• 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
• The feature is well documented and sample code can be built with the project CMake