[Bug] ConvertMask2BoxType function has the wrong order of height and width for PolygonMasks

[niecongchong]

Prerequisite

• I have searched Issues and Discussions but cannot get the expected help.
• I have read the FAQ documentation but cannot get the expected help.
• The bug has not been fixed in the latest version (master) or latest version (1.x).

Task

I have modified the scripts/configs, or I'm working on my own tasks/models/datasets.

Branch

1.x branch https://github.com/open-mmlab/mmrotate/tree/1.x

Environment

.

Reproduces the problem - code sample

.

Reproduces the problem - command or script

.

Reproduces the problem - error message

.

Additional information

ConvertMask2BoxType function has the wrong order of height and width for PolygonMasks, the developer should check and fix it.

    def transform(self, results: dict) -> dict:
        """The transform function."""
        assert 'gt_masks' in results.keys()
        masks = results['gt_masks']
        results['gt_bboxes'] = self.box_type_cls.from_instance_masks(masks)
        if not self.keep_mask:
            results.pop('gt_masks')

        # Modify results['instances'] for RotatedCocoMetric
        converted_instances = []
        for instance in results['instances']:
            m = np.array(instance['mask'][0])
            m = PolygonMasks([[m]], results['ori_shape'][1],
                             results['ori_shape'][0])     # here the order of height and width is wrong!!!!!!!!!!
            instance['bbox'] = self.box_type_cls.from_instance_masks(
                m).tensor[0].numpy().tolist()
            if not self.keep_mask:
                instance.pop('mask')
            converted_instances.append(instance)
        results['instances'] = converted_instances