image.c

Go to the documentation of this file.

/*
 * Copyright (C) 2015 Freek van Tienen <freek.v.tienen@gmail.com>
 *
 * This file is part of Paparazzi.
 *
 * Paparazzi is free software; you can redistribute it and/or modify
 * it under the terms of the GNU General Public License as published by
 * the Free Software Foundation; either version 2, or (at your option)
 * any later version.
 *
 * Paparazzi is distributed in the hope that it will be useful,
 * but WITHOUT ANY WARRANTY; without even the implied warranty of
 * MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE.  See the
 * GNU General Public License for more details.
 *
 * You should have received a copy of the GNU General Public License
 * along with Paparazzi; see the file COPYING.  If not, write to
 * the Free Software Foundation, 59 Temple Place - Suite 330,
 * Boston, MA 02111-1307, USA.
 */

#include "image.h"
#include <stdlib.h>
#include <string.h>
#include "lucas_kanade.h"

void image_create(struct image_t *img, uint16_t width, uint16_t height, enum image_type type)
{
  // Set the variables
  img->type = type;
  img->w = width;
  img->h = height;

  // Depending on the type the size differs
  if (type == IMAGE_YUV422) {
    img->buf_size = sizeof(uint8_t) * 2 * width * height;
  } else if (type == IMAGE_JPEG) {
    img->buf_size = sizeof(uint8_t) * 2 * width * height;  // At maximum quality this is enough
  } else if (type == IMAGE_GRADIENT) {
    img->buf_size = sizeof(int16_t) * width * height;
  } else {
    img->buf_size = sizeof(uint8_t) * width * height;
  }

  img->buf = malloc(img->buf_size);
}

void image_free(struct image_t *img)
{
  if (img->buf != NULL) {
    free(img->buf);
    img->buf = NULL;
  }
}

void image_copy(struct image_t *input, struct image_t *output)
{
  if (input->type != output->type) {
    return;
  }

  output->w = input->w;
  output->h = input->h;
  output->buf_size = input->buf_size;
  output->ts = input->ts;
  output->eulers = input->eulers;
  output->pprz_ts = input->pprz_ts;
  memcpy(output->buf, input->buf, input->buf_size);
}

void image_switch(struct image_t *a, struct image_t *b)
{
  /* Remember everything from image a */
  struct image_t old_a;
  memcpy(&old_a, a, sizeof(struct image_t));

  /* Copy everything from b to a */
  memcpy(a, b, sizeof(struct image_t));

  /* Copy everything from the remembered a to b */
  memcpy(b, &old_a, sizeof(struct image_t));
}

void image_to_grayscale(struct image_t *input, struct image_t *output)
{
  uint8_t *source = input->buf;
  uint8_t *dest = output->buf;
  source++;

  // Copy the creation timestamp (stays the same)
  output->ts = input->ts;
  output->eulers = input->eulers;
  output->pprz_ts = input->pprz_ts;

  // Copy the pixels
  for (int y = 0; y < output->h; y++) {
    for (int x = 0; x < output->w; x++) {
      if (output->type == IMAGE_YUV422) {
        *dest++ = 127;  // U / V
      }
      *dest++ = *source;    // Y
      source += 2;
    }
  }
}

uint16_t image_yuv422_colorfilt(struct image_t *input, struct image_t *output, uint8_t y_m, uint8_t y_M, uint8_t u_m,
                                uint8_t u_M, uint8_t v_m, uint8_t v_M)
{
  uint16_t cnt = 0;
  uint8_t *source = (uint8_t *)input->buf;
  uint8_t *dest = (uint8_t *)output->buf;

  // Copy the creation timestamp (stays the same)
  output->ts = input->ts;

  // Go trough all the pixels
  for (uint16_t y = 0; y < output->h; y++) {
    for (uint16_t x = 0; x < output->w; x += 2) {
      // Check if the color is inside the specified values
      if (
        (dest[1] >= y_m)
        && (dest[1] <= y_M)
        && (dest[0] >= u_m)
        && (dest[0] <= u_M)
        && (dest[2] >= v_m)
        && (dest[2] <= v_M)
      ) {
        cnt ++;
        // UYVY
        dest[0] = 64;        // U
        dest[1] = source[1];  // Y
        dest[2] = 255;        // V
        dest[3] = source[3];  // Y
      } else {
        // UYVY
        char u = source[0] - 127;
        u /= 4;
        dest[0] = 127;        // U
        dest[1] = source[1];  // Y
        u = source[2] - 127;
        u /= 4;
        dest[2] = 127;        // V
        dest[3] = source[3];  // Y
      }

      // Go to the next 2 pixels
      dest += 4;
      source += 4;
    }
  }
  return cnt;
}

int check_color_yuv422(struct image_t *im, int x, int y, uint8_t y_m, uint8_t y_M, uint8_t u_m, uint8_t u_M, uint8_t v_m, uint8_t v_M)
{
  // odd pixels are uy
  // even pixels are vy
  // adapt x, so that we always have u-channel in index 0:
  if (x % 2 == 1) { x--; }

  // Is the pixel inside the image?
  if (x < 0 || x >= im->w || y < 0 || y >= im->h) {
    return 0;
  }

  // Take the right place in the buffer:
  uint8_t *buf = im->buf;
  buf += 2 * (y * (im->w) + x); // each pixel has two bytes

  if (
    (buf[1] >= y_m)
    && (buf[1] <= y_M)
    && (buf[0] >= u_m)
    && (buf[0] <= u_M)
    && (buf[2] >= v_m)
    && (buf[2] <= v_M)
  ) {
    // the pixel passes:
    return 1;
  } else {
    // the pixel does not:
    return 0;
  }
}

void set_color_yuv422(struct image_t *im, int x, int y, uint8_t Y, uint8_t U, uint8_t V) {

  // odd pixels are uy
  // even pixels are vy
  // adapt x, so that we always have u-channel in index 0:
  if (x % 2 == 1) { x--; }

  // Is the pixel inside the image?
  if (x < 0 || x >= im->w || y < 0 || y >= im->h) {
    return;
  }

  // Take the right place in the buffer:
  uint8_t *buf = im->buf;
  buf += 2 * (y * (im->w) + x); // each pixel has two bytes

  buf[0] = U;
  buf[1] = Y;
  buf[2] = V;
  buf[3] = Y;
}


void image_yuv422_downsample(struct image_t *input, struct image_t *output, uint16_t downsample)
{
  uint8_t *source = input->buf;
  uint8_t *dest = output->buf;
  uint16_t pixelskip = (downsample - 1) * 2;

  // Copy the creation timestamp (stays the same)
  output->ts = input->ts;

  // Go trough all the pixels
  for (uint16_t y = 0; y < output->h; y++) {
    for (uint16_t x = 0; x < output->w; x += 2) {
      // YUYV
      *dest++ = *source++; // U
      *dest++ = *source++; // Y
      *dest++ = *source++; // V
      source += pixelskip;
      *dest++ = *source++; // Y
      source += pixelskip;
    }
    // read 1 in every 'downsample' rows, so skip (downsample-1) rows after reading the first
    source += (downsample - 1) * input->w * 2;
  }
}

void image_add_border(struct image_t *input, struct image_t *output, uint8_t border_size)
{
  // Create padded image based on input
  image_create(output, input->w + 2 * border_size, input->h + 2 * border_size, input->type);

  uint8_t *input_buf = (uint8_t *)input->buf;
  uint8_t *output_buf = (uint8_t *)output->buf;

  // Skip first `border_size` rows, iterate through next input->h rows
  for (uint16_t i = border_size; i != (output->h - border_size); i++) {

    // Mirror first `border_size` columns
    for (uint8_t j = 0; j != border_size; j++) {
      output_buf[i * output->w + (border_size - 1 - j)] = input_buf[(i - border_size) * input->w + j];
    }

    // Copy corresponding row values from input image
    memcpy(&output_buf[i * output->w + border_size], &input_buf[(i - border_size) * input->w], sizeof(uint8_t) * input->w);

    // Mirror last `border_size` columns
    for (uint8_t j = 0; j != border_size; j++) {
      output_buf[i * output->w + output->w - border_size + j] = output_buf[i * output->w + output->w - border_size - 1 - j];
    }
  }

  // Mirror first `border_size` and last `border_size` rows
  for (uint8_t i = 0; i != border_size; i++) {
    memcpy(&output_buf[(border_size - 1) * output->w - i * output->w], &output_buf[border_size * output->w + i * output->w],
           sizeof(uint8_t) * output->w);
    memcpy(&output_buf[(output->h - border_size) * output->w + i * output->w],
           &output_buf[(output->h - border_size - 1) * output->w - i * output->w], sizeof(uint8_t) * output->w);
  }
}

void pyramid_next_level(struct image_t *input, struct image_t *output, uint8_t border_size)
{
  // Create output image, new image size is half the size of input image without padding (border)
  image_create(output, (input->w + 1 - 2 * border_size) / 2, (input->h + 1 - 2 * border_size) / 2, input->type);

  uint8_t *input_buf = (uint8_t *)input->buf;
  uint8_t *output_buf = (uint8_t *)output->buf;

  uint16_t row, col; // coordinates of the central pixel; pixel being calculated in input matrix; center of filer matrix
  uint16_t w = input->w;
  int32_t sum = 0;

  for (uint16_t i = 0; i != output->h; i++) {

    for (uint16_t j = 0; j != output->w; j++) {
      row = border_size + 2 * i; // First skip border, then every second pixel
      col = border_size + 2 * j;

      sum =    39 * (input_buf[(row - 2) * w + (col - 2)] + input_buf[(row - 2) * w + (col + 2)] +
                     input_buf[(row + 2) * w + (col - 2)] + input_buf[(row + 2) * w + (col + 2)]);
      sum +=  156 * (input_buf[(row - 2) * w + (col - 1)] + input_buf[(row - 2) * w + (col + 1)] +
                     input_buf[(row - 1) * w + (col + 2)] + input_buf[(row + 1) * w + (col - 2)]
                     + input_buf[(row + 1) * w + (col + 2)] + input_buf[(row + 2) * w + (col - 1)] + input_buf[(row + 2) * w + (col + 1)] +
                     input_buf[(row - 1) * w + (col - 2)]);
      sum +=  234 * (input_buf[(row - 2) * w + (col)] + input_buf[(row) * w    + (col - 2)] +
                     input_buf[(row) * w    + (col + 2)] + input_buf[(row + 2) * w + (col)]);
      sum +=  625 * (input_buf[(row - 1) * w + (col - 1)] + input_buf[(row - 1) * w + (col + 1)] +
                     input_buf[(row + 1) * w + (col - 1)] + input_buf[(row + 1) * w + (col + 1)]);
      sum +=  938 * (input_buf[(row - 1) * w + (col)] + input_buf[(row) * w    + (col - 1)] +
                     input_buf[(row) * w    + (col + 1)] + input_buf[(row + 1) * w + (col)]);
      sum += 1406 * input_buf[(row) * w    + (col)];

      output_buf[i * output->w + j] = sum / 10000;
    }
  }
}


void pyramid_build(struct image_t *input, struct image_t *output_array, uint8_t pyr_level, uint16_t border_size)
{
  // Pad input image and save it as '0' pyramid level
  image_add_border(input, &output_array[0], border_size);

  // Temporary holds 'i' level version of original image to be padded and saved as 'i' pyramid level
  struct image_t temp;

  for (uint8_t i = 1; i != pyr_level + 1; i++) {
    pyramid_next_level(&output_array[i - 1], &temp, border_size);
    image_add_border(&temp, &output_array[i], border_size);
    image_free(&temp);
  }
}

void image_subpixel_window(struct image_t *input, struct image_t *output, struct point_t *center,
                           uint32_t subpixel_factor, uint8_t border_size)
{
  uint8_t *input_buf = (uint8_t *)input->buf;
  uint8_t *output_buf = (uint8_t *)output->buf;

  // Calculate the window size
  uint16_t half_window = output->w / 2;

  uint32_t subpixel_w = (input->w - 2) * subpixel_factor;
  uint32_t subpixel_h = (input->h - 2) * subpixel_factor;

  // Go through the whole window size in normal coordinates
  for (uint16_t i = 0; i < output->w; i++) {
    for (uint16_t j = 0; j < output->h; j++) {
      // Calculate the subpixel coordinate
      uint32_t x = center->x + border_size * subpixel_factor + (i - half_window) * subpixel_factor;
      uint32_t y = center->y + border_size * subpixel_factor + (j - half_window) * subpixel_factor;

      BoundUpper(x, subpixel_w);
      BoundUpper(y, subpixel_h);

      // Calculate the original pixel coordinate
      uint16_t orig_x = x / subpixel_factor;
      uint16_t orig_y = y / subpixel_factor;

      // Calculate top left (in subpixel coordinates)
      uint32_t tl_x = orig_x * subpixel_factor;
      uint32_t tl_y = orig_y * subpixel_factor;

      // Check if it is the top left pixel
      if (tl_x == x &&  tl_y == y) {
        output_buf[output->w * j + i] = input_buf[input->w * orig_y + orig_x];
      } else {
        // Calculate the difference from the top left
        uint32_t alpha_x = (x - tl_x);
        uint32_t alpha_y = (y - tl_y);

        // Blend from the 4 surrounding pixels
        uint32_t blend = (subpixel_factor - alpha_x) * (subpixel_factor - alpha_y) * input_buf[input->w * orig_y + orig_x];
        blend += alpha_x * (subpixel_factor - alpha_y) * input_buf[input->w * orig_y + (orig_x + 1)];
        blend += (subpixel_factor - alpha_x) * alpha_y * input_buf[input->w * (orig_y + 1) + orig_x];
        blend += alpha_x * alpha_y * input_buf[input->w * (orig_y + 1) + (orig_x + 1)];

        // Set the normalized pixel blend
        output_buf[output->w * j + i] = blend / (subpixel_factor * subpixel_factor);
      }
    }
  }
}

void image_gradients(struct image_t *input, struct image_t *dx, struct image_t *dy)
{
  // Fetch the buffers in the correct format
  uint8_t *input_buf = (uint8_t *)input->buf;
  int16_t *dx_buf = (int16_t *)dx->buf;
  int16_t *dy_buf = (int16_t *)dy->buf;

  // Go trough all pixels except the borders
  for (uint16_t x = 1; x < input->w - 1; x++) {
    for (uint16_t y = 1; y < input->h - 1; y++) {
      dx_buf[(y - 1)*dx->w + (x - 1)] = (int16_t)input_buf[y * input->w + x + 1] - (int16_t)input_buf[y * input->w + x - 1];
      dy_buf[(y - 1)*dy->w + (x - 1)] = (int16_t)input_buf[(y + 1) * input->w + x] - (int16_t)
                                        input_buf[(y - 1) * input->w + x];
    }
  }
}

void image_calculate_g(struct image_t *dx, struct image_t *dy, int32_t *g)
{
  int32_t sum_dxx = 0, sum_dxy = 0, sum_dyy = 0;

  // Fetch the buffers in the correct format
  int16_t *dx_buf = (int16_t *)dx->buf;
  int16_t *dy_buf = (int16_t *)dy->buf;

  // Calculate the different sums
  for (uint16_t x = 0; x < dx->w; x++) {
    for (uint16_t y = 0; y < dy->h; y++) {
      sum_dxx += ((int32_t)dx_buf[y * dx->w + x] * dx_buf[y * dx->w + x]);
      sum_dxy += ((int32_t)dx_buf[y * dx->w + x] * dy_buf[y * dy->w + x]);
      sum_dyy += ((int32_t)dy_buf[y * dy->w + x] * dy_buf[y * dy->w + x]);
    }
  }

  // output the G vector
  g[0] = sum_dxx / 255;
  g[1] = sum_dxy / 255;
  g[2] = g[1];
  g[3] = sum_dyy / 255;
}

uint32_t image_difference(struct image_t *img_a, struct image_t *img_b, struct image_t *diff)
{
  uint32_t sum_diff2 = 0;
  int16_t *diff_buf = NULL;

  // Fetch the buffers in the correct format
  uint8_t *img_a_buf = (uint8_t *)img_a->buf;
  uint8_t *img_b_buf = (uint8_t *)img_b->buf;

  // If we want the difference image back
  if (diff != NULL) {
    diff_buf = (int16_t *)diff->buf;
  }

  // Go trough the imagge pixels and calculate the difference
  for (uint16_t x = 0; x < img_b->w; x++) {
    for (uint16_t y = 0; y < img_b->h; y++) {
      int16_t diff_c = img_a_buf[(y + 1) * img_a->w + (x + 1)] - img_b_buf[y * img_b->w + x];
      sum_diff2 += diff_c * diff_c;

      // Set the difference image
      if (diff_buf != NULL) {
        diff_buf[y * diff->w + x] = diff_c;
      }
    }
  }

  return sum_diff2;
}

int32_t image_multiply(struct image_t *img_a, struct image_t *img_b, struct image_t *mult)
{
  int32_t sum = 0;
  int16_t *img_a_buf = (int16_t *)img_a->buf;
  int16_t *img_b_buf = (int16_t *)img_b->buf;
  int16_t *mult_buf = NULL;

  // When we want an output
  if (mult != NULL) {
    mult_buf = (int16_t *)mult->buf;
  }

  // Calculate the multiplication
  for (uint16_t x = 0; x < img_a->w; x++) {
    for (uint16_t y = 0; y < img_a->h; y++) {
      int32_t mult_c = img_a_buf[y * img_a->w + x] * img_b_buf[y * img_b->w + x];
      sum += mult_c;

      // Set the difference image
      if (mult_buf != NULL) {
        mult_buf[y * mult->w + x] = mult_c;
      }
    }
  }

  return sum;
}

void image_show_points(struct image_t *img, struct point_t *points, uint16_t points_cnt)
{
  uint8_t color[4];
  color[0] = 255;
  color[1] = 255;
  color[2] = 255;
  color[3] = 255;

  image_show_points_color(img, points, points_cnt, color);

}

void image_show_points_color(struct image_t *img, struct point_t *points, uint16_t points_cnt, uint8_t *color)
{
  uint8_t *img_buf = (uint8_t *)img->buf;
  uint8_t pixel_width = (img->type == IMAGE_YUV422) ? 2 : 1;

  int cross_hair = 1;
  int size_crosshair = 5;

  // Go trough all points and color them
  for (int i = 0; i < points_cnt; i++) {
    if (!cross_hair) {
      uint32_t idx = pixel_width * points[i].y * img->w + points[i].x * pixel_width;
      img_buf[idx] = 255;

      // YUV422 consists of 2 pixels
      if (img->type == IMAGE_YUV422) {
        idx++;
        img_buf[idx] = 255;
      }
    } else {
      image_draw_crosshair(img, &(points[i]), color, size_crosshair);
    }
  }
}

void image_show_flow(struct image_t *img, struct flow_t *vectors, uint16_t points_cnt, uint8_t subpixel_factor)
{
  static uint8_t color[4] = {255, 255, 255, 255};
  static uint8_t bad_color[4] = {0, 0, 0, 0};
  image_show_flow_color(img, vectors, points_cnt, subpixel_factor, color, bad_color);
}

void image_show_flow_color(struct image_t *img, struct flow_t *vectors, uint16_t points_cnt, uint8_t subpixel_factor,
                           const uint8_t *color, const uint8_t *bad_color)
{
  static int size_crosshair = 5;

  // Go through all the points
  for (uint16_t i = 0; i < points_cnt; i++) {
    // Draw a line from the original position with the flow vector
    struct point_t from = {
      .x = vectors[i].pos.x / subpixel_factor,
      .y = vectors[i].pos.y / subpixel_factor
    };
    struct point_t to = {
      .x = (vectors[i].pos.x + vectors[i].flow_x) / subpixel_factor,
      .y = (vectors[i].pos.y + vectors[i].flow_y) / subpixel_factor
    };

    if (vectors[i].error >= LARGE_FLOW_ERROR) {
      image_draw_crosshair(img, &to, bad_color, size_crosshair);
      image_draw_line_color(img, &from, &to, bad_color);
    } else {
      image_draw_crosshair(img, &to, color, size_crosshair);
      image_draw_line_color(img, &from, &to, color);
    }
  }
}
void image_gradient_pixel(struct image_t *img, struct point_t *loc, int method, int *dx, int *dy)
{
  // create the simple and sobel filter only once:

  int gradient_x, gradient_y, index;
  gradient_x = 0;
  gradient_y = 0;

  // get image buffer and take into account YUV vs. grayscale:
  uint8_t *img_buf = (uint8_t *)img->buf;
  uint8_t pixel_width = (img->type == IMAGE_YUV422) ? 2 : 1;
  uint8_t add_ind = pixel_width - 1;

  // check if all pixels will fall in the image:
  if (loc->x >= 1 && (loc->x + 1) < img->w && loc->y >= 1 && (loc->y + 1) < img->h) {
    if (method == 0) {

      // *************
      // Simple method
      // *************

      // dx:
      index = loc->y * img->w * pixel_width + (loc->x - 1) * pixel_width;
      gradient_x -= (int) img_buf[index + add_ind];
      index = loc->y * img->w * pixel_width + (loc->x + 1) * pixel_width;
      gradient_x += (int) img_buf[index + add_ind];
      // dy:
      index = (loc->y - 1) * img->w * pixel_width + loc->x * pixel_width;
      gradient_y -= (int) img_buf[index + add_ind];
      index = (loc->y + 1) * img->w * pixel_width + loc->x * pixel_width;
      gradient_y += (int) img_buf[index + add_ind];
    } else {

      // *****
      // Sobel
      // *****
      static int Sobel[9] = { -1, 0, 1, -2, 0, 2, -1, 0, 1};
      static int total_sobel = 8;

      int filt_ind_y = 0;
      int filt_ind_x;
      for (int x = -1; x <= 1; x++) {
        for (int y = -1; y <= 1; y++) {
          index = (loc->y + y) * img->w * pixel_width + (loc->x + x) * pixel_width;
          if (x != 0) {
            filt_ind_x = (x + 1) % 3 + (y + 1) * 3;
            gradient_x += Sobel[filt_ind_x] * (int) img_buf[index + add_ind];
          }
          if (y != 0) {
            gradient_y += Sobel[filt_ind_y] * (int) img_buf[index + add_ind];
          }
          filt_ind_y++;
        }
      }
      gradient_x /= total_sobel;
    }
  }

  // TODO: more efficient would be to use dx, dy directly:
  (*dx) = gradient_x;
  (*dy) = gradient_y;
}

void image_draw_rectangle(struct image_t *img, int x_min, int x_max, int y_min, int y_max, uint8_t *color)
{
  struct point_t from, to;

  // bottom from left to right:
  from.x = x_min;
  from.y = y_min;
  to.x = x_max;
  to.y = y_min;
  image_draw_line_color(img, &from, &to, color);

  // from bottom right to top right:
  from.x = x_max;
  from.y = y_min;
  to.x = x_max;
  to.y = y_max;
  image_draw_line_color(img, &from, &to, color);

  // from top right to top left:
  from.x = x_max;
  from.y = y_max;
  to.x = x_min;
  to.y = y_max;
  image_draw_line_color(img, &from, &to, color);

  // from top left to bottom left:
  from.x = x_min;
  from.y = y_max;
  to.x = x_min;
  to.y = y_min;
  image_draw_line_color(img, &from, &to, color);

}

void image_draw_crosshair(struct image_t *img, struct point_t *loc, const uint8_t *color, uint32_t size_crosshair)
{
  struct point_t from, to;

  if (loc->x >= size_crosshair && loc->x < img->w - size_crosshair
      && loc->y >= size_crosshair && loc->y < img->h - size_crosshair) {
    // draw the lines:
    from.x = loc->x - size_crosshair;
    from.y = loc->y;
    to.x = loc->x + size_crosshair;
    to.y = loc->y;
    image_draw_line_color(img, &from, &to, color);
    from.x = loc->x;
    from.y = loc->y - size_crosshair;
    to.x = loc->x;
    to.y = loc->y + size_crosshair;
    image_draw_line_color(img, &from, &to, color);
  }
}

void image_draw_line(struct image_t *img, struct point_t *from, struct point_t *to)
{
  static uint8_t color[4] = {255, 255, 255, 255};
  image_draw_line_color(img, from, to, color);
}


void image_draw_line_color(struct image_t *img, struct point_t *from, struct point_t *to, const uint8_t *color)
{
  int xerr = 0, yerr = 0;
  uint8_t *img_buf = (uint8_t *)img->buf;
  uint8_t pixel_width = (img->type == IMAGE_YUV422) ? 2 : 1;
  uint16_t startx = from->x;
  uint16_t starty = from->y;

  uint8_t temp_color[4] = {color[0], color[1], color[2], color[3]};

  /* compute the distances in both directions */
  int32_t delta_x = to->x - from->x;
  int32_t delta_y = to->y - from->y;

  /* Compute the direction of the increment,
     an increment of 0 means either a horizontal or vertical
     line.
  */
  int8_t incx, incy;
  if (delta_x > 0) { incx = 1; }
  else if (delta_x == 0) { incx = 0; }
  else { incx = -1; }

  if (delta_y > 0) { incy = 1; }
  else if (delta_y == 0) { incy = 0; }
  else { incy = -1; }

  /* determine which distance is greater */
  uint16_t distance = 0;
  delta_x = abs(delta_x);
  delta_y = abs(delta_y);
  if (delta_x > delta_y) { distance = delta_x * 20; }
  else { distance = delta_y * 20; }

  /* draw the line */
  for (uint16_t t = 0; /* starty >= 0 && */ starty < img->h && /* startx >= 0 && */ startx < img->w
       && t <= distance + 1; t++) {

    // depending on startx being odd or even, we first have to set U or V
    if (startx % 2 == 1) {
      temp_color[0] = color[2];
      temp_color[2] = color[0];
    } else {
      temp_color[0] = color[0];
      temp_color[2] = color[2];
    }
    uint32_t buf_loc = img->w * pixel_width * starty + startx * pixel_width;
    img_buf[buf_loc] = temp_color[0]; // u (when startx even)

    if (img->type == IMAGE_YUV422) {
      img_buf[buf_loc + 1] = temp_color[1]; // y1

      if (startx + 1 < img->w) {
        img_buf[buf_loc + 2] = temp_color[2]; // v (when startx even)
        img_buf[buf_loc + 3] = temp_color[3]; // y2
      }
    }

    xerr += delta_x;
    yerr += delta_y;
    if (xerr > distance) {
      xerr -= distance;
      startx += incx;
    }
    if (yerr > distance) {
      yerr -= distance;
      starty += incy;
    }
  }
}