Returns: deepcopy of current ObjectAnnotation instance

[xmin, ymin, xmax, ymax]

Returns the package version as a string and the package name as a string.

Loads json formatted data (given as "data") from load_path Encoding type can be specified with 'encoding' argument Example inputs:

Raise error if module is not installed.

Args: box (List[int]): [x1, y1, x2, y2]

Returns: [xmin, ymin, xmax, ymax]

Adds image to this Coco instance Args: image: CocoImage

Saves json formatted data (given as "data") as save_path Example inputs: data: {"image_id": 5} save_path: "dirname/coco.json"

Creates CocoCategory object using coco category list. Args: coco_category_list: List[Dict] [

Creates coco object from COCO formatted dict or COCO dataset file path. Args: coco_dict_or_path: dict/str or List[dict/s

Init Mask from coco segmentation representation. Args: segmentation : List[List] [ [

Loads a DetectionModel from given path. Args: model_type: str Name of the detection framework (example:

Function for performing prediction for given image using given detection_model. Arguments: image: str or np.ndarray Loca

Returns: shifted BoundingBox

Loads an image as PIL.Image.Image. Args: image : Can be image path or url (str), numpy image (np.ndarray) or PIL.Image

Returns boolean mask as coco segmentation: [ [x1, y1, x2, y2, x3, y3, ...], [x1, y1, x2, y2, x3, y3, ...],

Returns: [xmin, ymin, width, height]

Adds annotation to this CocoImage instance annotation : CocoAnnotation

Args: box1 (np.ndarray): np.array([x1, y1, x2, y2]) box2 (np.ndarray): np.array([x1, y1, x2, y2])

Creates image symlinks and annotation txts in yolo format from coco dataset. Args: output_dir: str Export directory.

Creates ObjectAnnotation from coco bbox [minx, miny, width, height] Args: bbox: List [minx, miny, width,

Args: res: resolution of image such as low, medium height: width: Returns: trigger slicing params function

Rearranges category mapping of given COCO dictionary based on given category_mapping. Can also be used to filter some of the categories.

Visualizes prediction category names, bounding boxes over the source image and exports it to output folder. Arguments: object_pre

Creates Coco object. Args: name: str Name of the Coco dataset, it determines exported json name.

According to image resolution calculation overlap params Args: orientation: image capture angle slide: sliding window

Generate voc bbox ([xmin, ymin, xmax, ymax]) from given bool_mask (2D np.ndarray)

Accepts shapely box/poly object and returns its bounding box in coco and voc formats

Convert coco segmentation to 2D boolean mask of given height and width

Function for slice image + get predicion for each slice + combine predictions in full image. Args: image: str or np.ndarray

This function should be implemented in a way that detection model should be initialized and set to self.model. (self.model_pa

Apply non-maximum suppression to avoid detecting too many overlapping bounding boxes for a given object. Args: predictions: (tens

[xmin, ymin, width, height]

Attach a text label above each bar in *rects*, displaying its height.

Adds category to this Coco instance Args: category: CocoCategory

Check whether annotation coordinates lie inside slice coordinates. Args: annotation (dict): Single annotation entry in COCO format.

Applies color mask to given input image.

Args: box1 (List[int]): [x1, y1, x2, y2] box2 (List[int]): [x1, y1, x2, y2]

Creates COCO dict with fields "images", "annotations", "categories". Arguments --------- images : List of CocoImage containing a

Creates CocoAnnotation object from category name and COCO formatted annotation dict (with fields "bbox", "segmentation", "category_id

Creates CocoAnnotation object using coco bbox Args: bbox: List [xmin, ymin, width, height] c

Creates CocoCategory object using coco category. Args: category: Dict {"supercategory": "person", "id":

Creates CocoImage object from COCO formatted image dict (with fields "id", "file_name", "height" and "weight"). Args: im

Accepts coco style polygon coords and converts it to shapely multipolygon object

Apply greedy version of non-maximum merging to avoid detecting too many overlapping bounding boxes for a given object. Args: obje

Combines 2 coco formatted annotations dicts, and returns the combined coco dict. Arguments: --------- coco_dict1 : dict

Apply non-maximum merging to avoid detecting too many overlapping bounding boxes for a given object. Args: object_predictions_as_

Slice a large image into smaller windows. If output_file_name is given export sliced images. Args: image (str or PIL.Image): File pat

Split images into train-val and returns them as sahi.utils.coco.Coco objects. Args: train_split_rate: float

Returns sahi.utils.coco.CocoPrediction representation of ObjectAnnotation.

[ [x1, y1, x2, y2, x3, y3, ...], [x1, y1, x2, y2, x3, y3, ...], ... ]

Arguments: score: prediction score between 0 and 1

Applies category remapping based on mapping given in self.category_remapping

This function should be implemented in a way that self._original_predictions should be converted to a list of prediction.ObjectPredic

Adds category to this CocoVid instance Args: category: CocoCategory

Cli app.

Apply greedy version of non-maximum merging per category to avoid detecting too many overlapping bounding boxes for a given object. Args:

Apply non-maximum merging per category to avoid detecting too many overlapping bounding boxes for a given object. Args: object_pr

Apply non-maximum suppression to avoid detecting too many overlapping bounding boxes for a given object. Args: predictions: (tens

Args: width: height: Returns: image capture orientation

According to image resolution calculate power(2,n) and return the closest smaller `n`. Args: resolution: the width and height of the

This function calculate according to image resolution slice and overlap params. Args: resolution: str height: int wid

Returns the ratio of intersection area to the smaller box's area

Returns the ratio of intersection area to the union

Iterates over all annotations and calculates total number of

Raise error if module version is not compatible.

Converts original predictions of the detection model to a list of prediction.ObjectPrediction object. Should be called after perform_