LiDAR Data Specification

LiDAR Data

By organizing LiDAR data into a single JSON file, you can quickly import all data required for reconstruction into MipMap Desktop with one click, or merge it with reconstruction parameters and pass it to the MipMap SDK for fully automated textured model and Gaussian splatting reconstruction.

For easy identification, change the JSON file extension to .mpl before importing into MipMap Desktop (for example, mipmap_lidar.mpl).

The data includes:

✅ Point cloud data: Typically LAS-format point clouds. Points must record acquisition timestamps in the gps_time field.
✅ Image data: Image set, including path, timestamp, position, and attitude angles (optional) for each image.
✅ Camera parameters: Focal length, principal point intrinsics, and distortion parameters.

LiDAR Data Fields

Field	Type	Required	Description
`coordinate_systems`	JSON	Yes	Coordinate system definitions. See coordinate_systems.
`camera_meta_data`	JSON	Yes	Camera parameters. See camera_meta_data.
`image_meta_data`	JSON	Yes	Image collection. See image_meta_data.
`lidar_data`	JSON	Yes	LiDAR data. See lidar_data.

coordinate_systems

coordinate_systems is an array of coordinate systems, each with an ID.

Field	Type	Required	Description
`coordinate_system`	JSON	Yes	Coordinate system object. See coordinate_system.
`id`	int	Yes	Coordinate system ID

The following example includes one local coordinate system and one WGS 84 geographic coordinate system:

json

"coordinate_systems": [
  {
    "id": 0,
    "coordinate_system": {
      "type": 1,
      "type_name": "Local"
    }
  },
  {
    "id": 1,
    "coordinate_system": {
      "type": 2,
      "label": "WGS 84",
      "epsg_code": 4326
    }
  }
]

camera_meta_data

Field	Type	Required	Description
`id`	int	Yes	Unique camera ID
`meta_data`	object	Yes	Camera metadata. See table below.

meta_data sub-fields:

Field	Type	Required	Description
`projection_model`	int	Yes	Lens model: `0` — perspective (k1, k2, p1, p2, k3); `1` — fisheye (k1, k2, k3, k4)
`camera_name`	string	Yes	Camera name
`width`	int	Yes	Sensor width (pixels)
`height`	int	Yes	Sensor height (pixels)
`parameters`	array	Yes	Intrinsic parameter array (perspective order: fx, fy, cx, cy, k1, k2, p1, p2, k3; fisheye order: fx, fy, cx, cy, k1, k2, k3, k4)
`constant_parameters`	array	No	Index array of intrinsics to hold fixed during bundle adjustment, in `parameters` order (0-based). For example, `[0,1,2,3,4,5,6,7,8]` for a perspective camera fixes all parameters.
`mask_path`	string	No	Path to a binary mask image (PNG). Pixels with value `0` are ignored during reconstruction (applied to all images from this camera).

TIP

A typical use case for camera masks: when a fixed camera on a robot captures parts of the robot body, use a mask to exclude those pixels from reconstruction.

Perspective lens

json

"camera_meta_data": [
  {
    "id": 1,
    "meta_data": {
      "projection_model": 0,    // perspective camera
      "camera_name": "common",
      "width": 5280,
      "height": 3956,
      "parameters": [
        3713.29,                // fx
        3713.29,                // fy
        2647.02,                // cx
        1969.28,                // cy
        -0.11257524,            // k1
        0.01487443,             // k2
        -0.00008572,            // p1
        1e-7,                   // p2
        -0.02706411             // k3
      ]
    }
  }
]

Fisheye lens

json

"camera_meta_data": [
  {
    "id": 1,
    "meta_data": {
      "camera_name": "fisheye",         // fisheye camera
      "projection_model": 1,
      "width": 3600,
      "height": 3600,
      "parameters": [
        982.7593599212141,              // fx
        982.7593599212141,              // fy
        1747.6373897301492,             // cx
        1806.4116030074354,             // cy
        0.03702410479839055,            // k1
        -0.016007338300982825,          // k2
        -1.0884582901480562e-05,        // k3
        -9.773097281093723e-05          // k4
      ]
    }
  }
]

image_meta_data

Field	Type	Required	Description
`id`	int	Yes	Unique image ID
`path`	string	Yes	Absolute image path
`meta_data`	object	Yes	Image metadata. See table below.
`crs_id`	int	No	Coordinate system ID for the image. Default: `0`.
`source_id`	int	Yes	Acquisition source ID. Default: `0`.

TIP

Use crs_id to distinguish image groups whose POS uses different coordinate systems—for example, WGS 84 or CGCS2000 for aerial data and a local CRS for ground data.

Use source_id to distinguish images from different acquisition sources—for example, UAV imagery versus handheld captures. Assigning different source_id values is essential.

Data sources in different coordinate systems must use different source IDs.

Typical use case: air–ground fusion.

meta_data sub-fields:

Field	Type	Required	Description
`width`	int	Yes	Image width (pixels)
`height`	int	Yes	Image height (pixels)
`camera_id`	int	Yes	Corresponding camera ID
`timestamp`	double	Yes	Image acquisition timestamp. Required for handheld LiDAR imagery; must match timestamps recorded in the LAS point cloud `gps_time` field. Optional for other sources.
`pos`	array[3]	Yes	POS position [longitude/X, latitude/Y, elevation/Z]. Required for handheld LiDAR imagery; must use the same CRS as the corresponding LAS point cloud.
`pos_sigma`	array[3]	Yes	POS accuracy. Centimeter-level RTK: `[0.03, 0.03, 0.06]`; decimeter-to-meter GNSS: `[0.5, 0.5, 1]`; meter-level and above: `[2.0, 2.0, 5.0]`. LiDAR data is recommended to use centimeter-level values.
`orientation`	array[9]	No	3×3 rotation matrix (9 values, row-major, unit vectors; world-to-camera). Strongly recommended for fisheye images.
`position_constant`	bool	No	Fix position (`true`/`false`). When fixed, bundle adjustment will not optimize image position.
`mask_path`	string	No	Path to a binary mask image (PNG). Pixels with value `0` are ignored during reconstruction.

Note: Binary masks mark invalid regions. The SDK ignores them during reconstruction to improve efficiency and produce cleaner outputs.

lidar_data

Field	Type	Description
`laser_meta_data`	array	LiDAR point cloud file metadata list

laser_meta_data

Field	Type	Description
`id`	int	LAS point cloud data ID
`path`	string	LAS point cloud file path (`.las` format)
`source_id`	int	`source_id` for the LAS data. Must match the paired images. Default: `0`.
`crs_id`	int	Coordinate system ID for the LAS data. Default: `0`.

Example:

json

{
  "camera_meta_data": [
  {
    "id": 1,
    "meta_data": {
      "camera_name": "Left",
      "projection_model": 1,
      "width": 3600,
      "height": 3600,
      "parameters": [
        982.7593599212141,
        982.7593599212141,
        1747.6373897301492,
        1806.4116030074354,
        0.03702410479839055,
        -0.016007338300982825,
        -1.0884582901480562e-05,
        -9.773097281093723e-05
      ]
    }
  },
  {
    "id": 2,
    "meta_data": {
      "camera_name": "Right",
      "projection_model": 1,
      "width": 3600,
      "height": 3600,
      "parameters": [
        982.7593599212141,
        982.7593599212141,
        1747.6373897301492,
        1806.4116030074354,
        0.03702410479839055,
        -0.016007338300982825,
        -1.0884582901480562e-05,
        -9.773097281093723e-05
      ]
    }
  }
  ],
  "image_meta_data":[
    {
        "id": 1,
        "path": "\\left\\2390-195042-472_00395-L.JPG",
        "crs_id": 0,
        "source_id": 0,
        "meta_data": {
        "width": 3600,
        "height": 3600,
        "camera_id": 1,
        "timestamp": 1.0000,
        "pos": [
            -2272303.077314121,
            5011055.355739306,
            3215292.943560472
        ],
        "pos_sigma": [0.03,0.03,0.06],
        "orientation": [
            0.8929275412022056,
            0.43070972182403255,
            0.13103259781004117,
            -0.4500241512548592,
            0.8457917052937085,
            0.2865565468519331,
            0.012596406230632032,
            -0.31484206641207435,
            0.9490604847782084
        ]
        }
    },
    {
        "id": 2,
        "path": "\\right\\2390-195042-472_00395-R.JPG",
        "crs_id": 0,
        "source_id": 0,
        "meta_data": {
        "width": 3600,
        "height": 3600,
        "camera_id": 2,
        "timestamp": 2.0000,
        "pos": [
            -2272303.117693735,
            5011055.368901664,
            3215292.967908075
        ],
        "pos_sigma": [0.03,0.03,0.06],
        "orientation": [
            0.8918935649712837,
            0.435497387453275,
            0.12193397510204676,
            -0.4504120798442928,
            0.8311302214603127,
            0.3261157973874772,
            0.040679566039387514,
            -0.34578111645476495,
            0.9374329802231962
        ]
        }
    },
    {
        "id": 2,
        "path": "\\DJI_0001.JPG",
        "crs_id": 1,
        "source_id": 1,
        "meta_data": {
        "width": 3600,
        "height": 3600,
        "camera_id": 2,
        "timestamp": 2.0000,
        "pos": [
            -2272303.117693735,
            5011055.368901664,
            3215292.967908075
        ],
        "pos_sigma": [0.03,0.03,0.06],
        "orientation": [
            0.8918935649712837,
            0.435497387453275,
            0.12193397510204676,
            -0.4504120798442928,
            0.8311302214603127,
            0.3261157973874772,
            0.040679566039387514,
            -0.34578111645476495,
            0.9374329802231962
        ]
        }
    }
  ],
  "lidar_data": {
    "laser_meta_data": [
      {
        "id": 1,
        "path": "C:/path/to/file1.las",
        "source_id": 0,
        "crs_id": 0
      },
      {
        "id": 2,
        "path": "C:/path/to/file2.las",
        "source_id": 0,
        "crs_id": 0
      }
    ]
  },
  "coordinate_systems": [
    {
      "id": 0,
      "coordinate_system": {
        "type": 1,
        "type_name": "Local"
      }
    },
    {
      "id": 1,
      "coordinate_system": {
        "type": 2,
        "type_name": "Geographic",
        "label": "WGS 84",
        "epsg_code": 4326
      }
    }
  ]
}

LiDAR Data ​

LiDAR Data Fields ​

coordinate_systems ​

camera_meta_data ​

image_meta_data ​

lidar_data ​

laser_meta_data ​

LiDAR Data

LiDAR Data Fields

coordinate_systems

camera_meta_data

image_meta_data

lidar_data

laser_meta_data