image.h

/* Copyright (c) 2017 VisualDL Authors. All Rights Reserve.

Licensed under the Apache License, Version 2.0 (the "License");
you may not use this file except in compliance with the License.
You may obtain a copy of the License at

    http://www.apache.org/licenses/LICENSE-2.0

Unless required by applicable law or agreed to in writing, software
distributed under the License is distributed on an "AS IS" BASIS,
WITHOUT WARRANTIES OR CONDITIONS OF ANY KIND, either express or implied.
See the License for the specific language governing permissions and
limitations under the License. */

#ifndef VISUALDL_UTILS_IMAGE_H
#define VISUALDL_UTILS_IMAGE_H

#include <Eigen/Core>
#include <algorithm>
#include <unsupported/Eigen/CXX11/Tensor>
#include "visualdl/utils/logging.h"

namespace visualdl {

using uint8_t = unsigned char;

/*
 * 2: height*width, channel
 */
template <typename T>
using ImageDT =
    Eigen::Matrix<T, Eigen::Dynamic, Eigen::Dynamic, Eigen::RowMajor>;
using Uint8Image = ImageDT<uint8_t>;

/*
 * hw: height*width
 * depth: number of channels
 */
static void NormalizeImage(Uint8Image* image,
                           const float* buffer,
                           int hw,
                           int depth) {
  // Both image and buffer should be used in row major.
  Eigen::Map<const Eigen::
                 Matrix<float, Eigen::Dynamic, Eigen::Dynamic, Eigen::RowMajor>>
      values(buffer, depth, hw);

  CHECK_EQ(image->size(), hw * depth);
  CHECK_EQ(image->row(0).size(), hw);
  CHECK_EQ(image->col(0).size(), depth);

  std::vector<int> infinite_pixels;
  // compute min and max ignoring nonfinite pixels
  float image_min = std::numeric_limits<float>::infinity();
  float image_max = -image_min;
  for (int i = 0; i < hw; i++) {
    bool finite = true;
    for (int j = 0; j < depth; j++) {
      // if infinite, skip this pixel
      if (!std::isfinite(values(j, i))) {
        infinite_pixels.emplace_back(i);
        finite = false;
        break;
      }
    }
    if (finite) {
      for (int j = 0; j < depth; j++) {
        float v = values(j, i);
        image_min = std::min(image_min, v);
        image_max = std::max(image_max, v);
      }
    }
  }

  // Pick an affine transform into uint8
  const float kZeroThreshold = 1e-6;
  float scale, offset;
  if (image_min < 0) {
    float max_val = std::max(std::abs(image_min), image_max);
    scale = (max_val < kZeroThreshold ? 0.0f : 127.0f) / max_val;
  } else {
    scale = (image_max < kZeroThreshold ? 0.0f : 255.0f) / image_max;
    offset = 0.0f;
  }

  // Transform image, turning nonfinite values to bad_color
  for (int i = 0; i < depth; i++) {
    auto tmp = scale * values.row(i).array() + offset;
    image->row(i) = tmp.cast<uint8_t>();
  }

  for (int pixel : infinite_pixels) {
    for (int i = 0; i < depth; i++) {
      // TODO(ChunweiYan) use some highlight color to represent infinite pixels.
      (*image)(pixel, i) = (uint8_t)0;
    }
  }
}

}  // namespace visualdl

#endif