+ +

The (Circle) Circular Camera Sample Overview

+ +

The (Circle) Circular Camera sample application demonstrates how you can handle Camera API, among others how to start camera preview and capture images.

+ +

The following figure illustrates two screens of the (Circle) Circular Camera sample application: the camera preview window and the camera attributes menu views.

+ +

Figure 1: (Circle) Circular camera views

(Circle) Circular camera views

+ +

Click the red camera button to capture a picture or press the preview to enter the camera attributes list.

From there you can set any camera attributes this application makes available for you.

+ +

Figure 2: (Circle) Circular camera UI structure

(Circle) Circular camera UI structure +

+ +

Prerequisites

To ensure proper application execution, the http://tizen.org/privilege/mediastorage and +the http://tizen.org/privilege/camera privileges +must be set. Also, there should be the camera device installed on your device.

+ +

Implementation

+ +

View module

The most important part of the view module is the camera preview surface creation. For this purpose you can use the +evas filled image.

+ +

+Evas_Object *preview_surface_create(Evas_Object *parent, previev_surface_long_press_cb cb)
+{
+  Evas_Object *surface = evas_object_image_filled_add(evas_object_evas_get(parent));
+	// Common initialization code
+  return surface;
+}
+

+ +

Controller module

The cam module is the essential part of this application. It allows for the various camera attributes to be set.

The initialization function takes the preview surface (shown in view section) as a parameter. It creates a camera handle and +adds callbacks for the following events: +

state change,
operation interruption,
error.

+ +

+bool cam_init(const void *evas_display_surface)
+{
+	// Common initialization code
+
+  cmd.evas_display_surface = (void *)evas_display_surface;
+
+  ok = (CAMERA_ERROR_NONE == camera_create(CAMERA_DEVICE_CAMERA0, &cmd.camera));
+  ok &= (CAMERA_ERROR_NONE == camera_set_state_changed_cb(cmd.camera, __camera_state_changed_cb, NULL));
+  ok &= (CAMERA_ERROR_NONE == camera_set_interrupted_cb(cmd.camera, __camera_interrupted_cb, NULL));
+  ok &= (CAMERA_ERROR_NONE == camera_set_error_cb(cmd.camera, __camera_error_cb, NULL));
+
+  camera_get_capture_format(cmd.camera, &cmd.pixel_format);
+
+  return ok;
+}
+

+static void __camera_state_changed_cb(camera_state_e previous, camera_state_e current,
+											bool by_policy, void *user_data)
+{
+  dlog_print(DLOG_INFO, LOG_TAG, "Camera state has changed from: %d to: %d by policy: %d", previous, current, by_policy);
+}
+

+static void __camera_interrupted_cb(camera_policy_e policy, camera_state_e previous,
+										camera_state_e current, void *user_data)
+{
+  dlog_print(DLOG_INFO, LOG_TAG, "Camera state interrupted. State has changed from: %d to: %d by policy: %d",
+        previous, current, policy);
+}
+

It is important to free the camera handle and release its resources when an error occurs.

+static void __camera_error_cb(camera_error_e error, camera_state_e current_state, void *user_data)
+{
+  dlog_print(DLOG_ERROR, LOG_TAG, "Camera error: %s has occurred; current state: %d, deinitializing",
+        get_error_message(error), current_state);
+  cam_deinit();
+}
+

Once the initialization is succesfully finished, the preview function can be called. Please note that +the camera state is always checked at the beginning. Starting the preview will fail unless the camera is in one of the following states: +

CAMERA_STATE_CREATED,
CAMERA_STATE_CAPTURED.

+You might also notice how the preview surface (previously passed as an argument to the initialization function) is used in +camera_set_display() +function. The GET_DISPLAY() API macro has to be used as it is a portable way to convert the +Evas_Object * +to the +camera_display_h +.

+bool cam_start_preview(void)
+{
+  camera_state_e state = CAMERA_STATE_NONE;
+  int width = 0;
+  int height = 0;
+
+  camera_get_state(cmd.camera, &state);
+
+  if (state == CAMERA_STATE_PREVIEW)
+    return true;
+
+  if (state != CAMERA_STATE_CREATED && state != CAMERA_STATE_CAPTURED)
+    return false;
+
+  if (CAMERA_ERROR_NONE == camera_get_recommended_preview_resolution(cmd.camera, &width, &height)) {
+    camera_set_preview_resolution(cmd.camera, width, height);
+  } else {
+    return false;
+  }
+
+  camera_set_display(cmd.camera, CAMERA_DISPLAY_TYPE_EVAS, GET_DISPLAY(cmd.evas_display_surface));
+  camera_set_display_mode(cmd.camera, CAMERA_DISPLAY_MODE_FULL);
+  camera_set_display_visible(cmd.camera, true);
+  camera_start_focusing(cmd.camera, true);
+
+  return (CAMERA_ERROR_NONE == camera_start_preview(cmd.camera));
+}
+

+ +

Once the preview is successfully set up, the camera view is visible on the screen and it is possible to capture photos.

Photo taking is implemented in cam_start_capturing(), +__camera_capturing_cb() and +__camera_capture_completed_cb() functions.

+ +

+bool cam_start_capturing(void)
+{
+  camera_state_e state = CAMERA_STATE_NONE;
+
+  if (CAMERA_PIXEL_FORMAT_INVALID == cmd.pixel_format) {
+    dlog_print(DLOG_ERROR, LOG_TAG, "Error - capture format hasn't been selected");
+    return false;
+  }
+
+  camera_get_state(cmd.camera, &state);
+
+  if (state == CAMERA_STATE_CAPTURING)
+    return true;
+
+  if (state == CAMERA_STATE_NONE || (state != CAMERA_STATE_PREVIEW && !cam_start_preview()))
+    return false;
+
+  if (!max_width || !max_height)
+    camera_foreach_supported_capture_resolution(cmd.camera, __supported_capture_resolution_cb, NULL);
+
+  if (max_width && max_height)
+    camera_set_capture_resolution(cmd.camera, max_width, max_height);
+
+  return (CAMERA_ERROR_NONE == camera_start_capture(cmd.camera, __camera_capturing_cb, __camera_capture_completed_cb, NULL));
+}
+

Once the camera device has filled the image buffer, the +__camera_capturing_cb() +callback is called. This is the place to save the captured photo. The +camera_image_data_s +structure contains a raw image data buffer and information about its dimensions and buffer length. +This application allows the user to select between different types of camera images format. One of them is +CAMERA_PIXEL_FORMAT_JPEG which is already encoded. +The rest of the formats are raw types. This is why jpeg format is distinguished in the following code.

+static void __camera_capturing_cb(camera_image_data_s *image, camera_image_data_s *postview,
+                    camera_image_data_s *thumbnail, void *user_data)
+{
+  FILE *f = NULL;
+  int w = 0;
+  char currently_saved_file[PATH_MAX] = {0,};
+
+  if (cmd.pixel_format == CAMERA_PIXEL_FORMAT_INVALID || !image)
+    return;
+
+  if (image->format == CAMERA_PIXEL_FORMAT_JPEG) {
+    snprintf(currently_saved_file, PATH_MAX, SAVE_FILE_PATH_PATTERN, time(NULL));
+    f = fopen(currently_saved_file, "w+");
+    if (!f) {
+      dlog_print(DLOG_ERROR, LOG_TAG, "Could not obtain handle for file %s ", currently_saved_file);
+      return;
+    }
+
+    do {
+      w += fwrite(image->data + w, sizeof(unsigned char), image->size - w, f);
+    } while (w < image->size);
+
+    fclose(f);
+  } else {
+    __encode_to_jpeg(image->data, image->width, image->height, currently_saved_file);
+  }
+}
+

+ +

When the capture callback returns and the camera's internal state is updated, another callback is invoked - +__camera_capture_completed_cb(). This +gives the developer a chance to change the camera state from CAMERA_STATE_CAPTURED +to +CAMERA_STATE_PREVIEW and allows the camera +view to be continuously displayed again.

+static void __camera_capture_completed_cb(void *user_data)
+{
+  cam_start_preview();
+}
+

+ + + +