Parameters of DEEP-GL

You are currently viewing the documentation for the latest version (2.4.1). To access a different version, click the "Switch version" button located in the upper-right corner of the page.

■ If you are not sure which version of the product you are currently using, please feel free to contact Mech-Mind Technical Support.

This topic introduces the parameters of DEEP-GL. The parameters are divided into parameters for the 2D image and parameters for the depth map / point cloud, based on the data types that they affect.

Parameters for the 2D Image

DEEP-GL provides two types of 2D images: 2D image (texture) and 2D images (depth source). The usage of each type is as follows:

2D image type Usage

2D image (texture)

Texture the point cloud

2D images (depth source)

Check intrinsic parameters

Set ROI

Perform hand-eye calibration

The 2D image should not be too bright or too dark, and the surface features of the target object should be clearly visible in the image.

The quality of the 2D image is affected by parameters in the 2D Parameters category and the Gain parameter in the 3D Parameters category.

If you are using a color camera, and the color of the image differ significantly from that of the objects, you can adjust the white balance setting. For detailed instructions, please refer to Adjust White Balance.

2D Parameters

2D Image (Texture) Exposure Mode

Description

Sets the exposure mode for capturing the 2D image (texture).

Visibility

Beginner, Expert, Guru

Values

  • Timed: Sets a single exposure time. Usually used in stable lighting conditions.

  • Auto: The exposure time is automatically adjusted. Usually used in varying lighting conditions.

  • HDR: Sets multiple exposure times and merge the images. Usually used for objects with various colors or textures.

Instruction

After selecting different options, different parameters are displayed in the 2D Parameters category for adjustment:

Timed: Exposure Time

Description

This parameter affects the brightness of the image.

  • Long exposure time increases the image brightness.

  • Short exposure time decreases the image brightness.

Visibility

Beginner, Expert, Guru

Values

0.1 to 999 ms

Instructions

Adjust this parameter based on the quality of the 2D image. The 2D image should not be too bright or too dark, and the surface features of the target object should be clearly visible in the image.

  • Long exposure time is usually used in dark environments.

  • Short exposure time is usually used in bright environments.

2D images obtained with different Exposure Time values (all other conditions identical):

5 ms 90 ms 500 ms

exposure time 5

exposure time 90

exposure time 500

Auto: Gray Value

Description

This parameter affects the brightness of the 2D image. Increase the value if the 2D image is too dark and decrease if too bright.

Visibility

Beginner, Expert, Guru

Values

0 to 255

Instruction

N/A

2D images obtained with different Gray Value values (all other conditions identical):

50 100 150

gray value 50

gray value 100

gray value 150
The Gray Value of a monochrome image is equivalent to the image brightness; the Gray Value of a color image is equivalent to the brightness of each color channel.

Auto: Auto-Exposure ROI

Description

  • The exposure time is automatically adjusted based on the lighting, object colors, etc., in the selected region.

  • When an auto-exposure ROI is not set, the exposure time is automatically adjusted based on the entire camera FOV.

Visibility

Beginner, Expert, Guru

Values

N/A

Instruction

For detailed instructions, refer to Set Auto-Exposure ROI below.

2D images obtained with different Auto-Exposure ROI values (all other conditions identical):

Auto-exposure ROI 2D image

auto roi 1

auto roi scan 1

auto roi 2

auto roi scan 2

auto roi 3

auto roi scan 3

Set Auto-Exposure ROI

  1. Click Edit to the right of Auto-Exposure ROI to open the Set ROI window.

  2. Select and adjust the ROI on the left. Drag the selection box to move it. Drag the handles on the box to adjust the box size.

  3. Click Apply to apply the set auto-exposure ROI.

    roi 2d
    Click Clear to remove the current auto-exposure ROI.

  4. Acquire data again and view the 2D image to check the result of automatic exposure adjustment.

HDR: Tone Mapping

Description

This function can make the image look more natural. If the 2D image appears very different from the actual objects, please enable this function.

Visibility

Beginner, Expert, Guru

Values

Tone Mapping toggle switch:

  • When the toggle switch is turned on, Tone Mapping will be applied.

  • When the toggle switch is turned off, Tone Mapping will not be applied.

2D images obtained with Tone Mapping off or on (all other conditions identical):

Off On

tone mapping false

tone mapping true

HDR: Exposure Sequence

Description

Set multiple exposure times, and the captured images are merged to generate a 2D image that retains more details in the highlights and shadows.

Visibility

Beginner, Expert, Guru

Values

N/A

Instructions

  1. Click Edit to the right of Exposure Sequence to open the editing window of the exposure sequence.

  2. Click + to add and set a new exposure time.

  3. To delete an exposure time, select it and click -.

  4. After you finish adding the exposure times, click Apply to apply the exposure sequence.

    • Cancel: do not save the current settings and exit this window.

    • Reset: clear all the exposure times.

2D images captured with single exposure times (all other conditions identical):

500 ms 700 ms 900 ms

hdr false 500ms

hdr false 700ms

hdr false 900ms

2D images captured with different exposure sequences consisting of the above exposure times (all other conditions identical):

700 ms 500ms & 900ms 500 ms & 700 ms & 900 ms

hdr false 1

hdr false 2

hdr false 3

2D Image (Depth Source) Exposure Mode

Description

Sets the exposure mode for capturing the 2D images (depth source).

Visibility

Beginner, Expert, Guru

Values

  • Timed: Sets a single exposure time. Usually used in stable lighting conditions.

  • Flash: Uses the projector for supplemental light. Usually used in dark environments.

Instruction

After selecting different options, different parameters are displayed in the 2D Parameters category for adjustment:
When using a DEEP-GL camera with Mech-Vision, the “Color Image” port in Mech-Vision outputs the 2D image (depth source). If you need to adjust the “Color Image” in Mech-Vision, please adjust 2D Image (Depth Source) Exposure Mode.

Timed: 2D Image (Depth Source) Exposure Time

Description

This parameter affects the brightness of the image.

  • Long exposure time increases the image brightness.

  • Short exposure time decreases the image brightness.

Visibility

Beginner, Expert, Guru

Values

0.1 to 999 ms

Instructions

Adjust this parameter based on the quality of the 2D image. The 2D image should not be too bright or too dark, and the surface features of the target object should be clearly visible in the image.

  • Long exposure time is usually used in dark environments.

  • Short exposure time is usually used in bright environments.

2D images obtained with different 2D Image (Depth Source) Exposure Time values (all other conditions identical):

10 ms 40 ms 80 ms

lsr timed 10

lsr timed 40

lsr timed 80

Flash: Flash Acquisition Mode

Description

Selects the mode for capturing 2D images when adding supplemental light with the projector. For detailed descriptions, refer to 2D Flash Exposure Mode.

Visibility

Expert, Guru

Values

  • Fast: The 2D image is acquired as part of the 3D data, thus providing a faster acquisition speed. However, the refresh rate of the 2D image during continuous acquisition is lower (every 5 seconds).

  • Responsive: The 2D image is acquired independently, which reduces the acquisition speed. However, the refresh rate of the 2D image during continuous acquisition is higher (every second).

Instruction

Adjust this parameter based on the actual needs for the data acquisition speed and the refresh rate of the 2D image during continuous acquisition.

When Responsive is selected, Flash Exposure Time will be displayed for adjustment.

  • When Fast is selected, the brightness of the 2D image is affected by the Exposure Time and Laser Power parameters in the 3D Parameters category.

  • When Responsive is selected, the brightness of the 2D image is affected by Laser Power parameter in the 3D Parameters category.

Flash: Flash Exposure Time

Description

Sets the exposure time for capturing 2D images when the value of Flash Acquisition Mode is set to Responsive.

Visibility

Beginner, Expert, Guru

Values

0.1–99ms

Instruction

Long exposure time is usually used in dark environments, and short exposure time is usually used in bright environment.

The value of this parameter for DEEP-GL and LSR must be a multiple of 4, and the minimum value that can be set is 8 ms. The entered value is automatically adjusted.

3D Parameters

Gain

Description

This parameter affects the brightness of the captured image. At the same time, more noise is introduced.

This parameter affects the quality of the depth map and point cloud by changing the brightness of the images used for calculating depth data.

Visibility

Expert, Guru

Values

0 to 16 dB

Instructions

When the expected brightness cannot be reached by adjusting Exposure Time, adjust this parameter.

2D images and depth maps obtained with different Gain values (all other conditions identical):

0 5 10

gain 0 2d

gain 5 2d

gain 10 2d

gain 0 depth

gain 5 depth

gain 10 depth

Parameters for the Depth Map / Point Cloud

In the depth map and point cloud, the needed data should be complete.

Determine which part of the data is needed according to your actual needs. For example, if you need to pick metal bowls with opening facing up by the rim, usually you only need to make sure that the data of the bowl rims is complete.

Parameters in the following categories affect the quality of the depth map and point cloud.

Parameter category Depth map Point cloud

3D Parameters

icon_yes

icon_yes

Point Cloud Processing

icon_yes

Depth Range

icon_yes

icon_yes

ROI

icon_yes

icon_yes

3D Parameters

The parameters in this category affect the images used for calculating depth data, thus affecting the quality of the depth map and point cloud.

Use Exposure Assistant can obtain the recommended exposure parameter values. Click Auto Set to the right of 3D Parameters to open Exposure Assistant.

Exposure Multiplier

Description

Set the number of Exposure Time.

Visibility

Beginner, Expert, Guru

Values

1 to 3

Instruction

  • When the value of Exposure Multiplier is greater than 1, multiple Exposure Time parameters must be set.

  • All images captured with different exposure times are used for calculating depth. Using multiple exposure times can improve the completeness of depth data but also increases processing time.

  • With multiple exposure times, the amount of time for obtaining the depth map and point cloud increases. Please try to minimize the number of exposure times while ensuring the data quality.

For detailed explanations, refer to Adjust 3D Exposure Settings.

Exposure Multiplier is unavailable when Fringe Coding Mode in the 3D Parameters category is set to Reflective.

Exposure Time

Description

Set the exposure time for acquiring depth information. The number of Exposure Time to be set depends on the value of Exposure Multiplier.

Visibility

Beginner, Expert, Guru

Values

  • Value range: 0.1 ms to 99 ms

Instructions

  • Usually, long exposure time is used for dark objects, and short exposure time is used for light objects.

  • Exposure times too long or too short both lead to loss of depth information.

  • The exposure time of a laser camera must be a multiple of 4. If the entered value is not a multiple of 4, it will be automatically adjusted. The minimum value that can be set is 4 ms for the Laser series and 8 ms for other laser cameras.

For detailed instructions, refer to Adjust 3D Exposure Settings.

Point clouds obtained under different Exposure Time values (all other conditions identical):

Exposure Time 1 ms Exposure Time 4 ms Exposure Time 10 ms Exposure Time 50 ms

exposure time 3d 1

exposure time 3d 4

exposure time 3d 10

exposure time 3d 50
The black regions lack depth information to generate corresponding point cloud.

Gain

Description

This parameter affects the brightness of the captured image. At the same time, more noise is introduced.

This parameter affects the quality of the depth map and point cloud by changing the brightness of the images used for calculating depth data.

Visibility

Expert, Guru

Values

0 to 16 dB

Instructions

When the expected brightness cannot be reached by adjusting Exposure Time, adjust this parameter.

2D images and depth maps obtained with different Gain values (all other conditions identical):

0 5 10

gain 0 2d

gain 5 2d

gain 10 2d

gain 0 depth

gain 5 depth

gain 10 depth

Laser

Fringe Coding Mode

Description

Select the pattern of the structured light to be projected.

Visibility

Expert, Guru

Values

  • Fast: suitable for non-reflective objects, provides fast acquisition speed, but the depth data quality is lower.

  • Accurate: suitable for non-reflective objects, provides high-quality depth data, but the acquisition speed is slower.

  • Reflective: suitable for reflective objects, provides high-quality depth data, but the acquisition speed is slower.

Instructions

Adjust this parameter according to the type of target objects and your actual needs for data quality and data acquisition speed.

When Reflective is selected, The Processing Mode parameter is displayed.

  • When Reflective is selected, the following tool and parameters are unavailable:

  • When the target object is non-reflective, the point clouds obtained with Fringe Coding Mode set to Accurate and Fast are as follows (all other conditions identical):

    Accurate Fast

    coding accurate

    coding fast

  • When the target object is reflective, the point clouds obtained with Fringe Coding Mode set to Accurate and Reflective are as follows (all other conditions identical):

    Accurate Reflective

    fringe for reflect accurate

    fringe for reflect reflective
Reflective: Processing Mode

Description

Select the data processing mode for the Reflective fringe coding mode.

Visibility

Expert, Guru

Values

  • Faster: provides faster processing speed, but the depth data might have missing points. Suitable for scenarios with relatively simple reflective conditions.

  • More complete: provides more complete depth data, but the processing speed is slower. Suitable for scenarios with complex reflective conditions, such as a bin whose walls often cause interreflection.

Instruction

Adjust this parameter according to your actual needs for data quality and data acquisition speed.

Point clouds obtained with different Processing Mode values (all other conditions identical):

Faster More complete

reflective faster 1

reflective more complete 1

reflective faster 2

reflective more complete 2
Laser Power

Description

Set the power of the laser projector, which affects the brightness of the projected structured light.

Visibility

Expert, Guru

Values

50 to 100%

Instruction

Usually, using the default value is sufficient.

  • Greater value results in higher structured light brightness; smaller value results in lower structured light brightness.

  • Greater values are suitable for dark objects, and smaller values are suitable for reflective objects.

Point Cloud Processing

Adjusting the parameters in the Point Cloud Processing category can improve the quality of the point cloud.

Principles for Adjusting Point Cloud Processing Parameters

When adjusting the Point Cloud Processing parameters, follow these principles to reduce the camera acquisition time and cycle time:

  1. Adjust Outlier Removal first. The calculation time for different intensities of this parameter is about the same. Using a higher intensity will not add to the calculation time too much.

  2. It is recommended to use low intensities of Surface Smoothing and Noise Removal. The higher the intensities of these two parameters, the longer the calculation time.

Surface Smoothing

Description

This parameter reduces the depth fluctuation in the point cloud and improve its resemblance to the actual object surface.

Visibility

Beginner, Expert, Guru

Values

  • Off

  • Weak

  • Normal

  • Strong

Instructions

  • Surface Smoothing causes loss of object surface details. Higher intensity causes more object details to be lost. Lower intensity causes fewer object details to be lost.

  • Higher intensity results in slower calculation, and lower intensity results in faster calculation.

Point clouds obtained with different Surface Smoothing values (all other conditions identical):

Off Weak

cloud smoothing off

cloud smoothing weak

Normal

Strong

cloud smoothing normal

cloud smoothing strong

Outlier Removal

Description

This parameter removes the outliers in the point cloud. Outliers are clustered points away from the object point cloud.

Visibility

Beginner, Expert, Guru

Values

  • Off

  • Weak

  • Normal

  • Strong

Instructions

  • Higher intensity removes more outliers, and lower intensity removes fewer outliers.

  • If the object point cloud contains multiple parts, high intensities of outlier removal might remove some parts. For example, if the object is a cup, outlier removal might remove the point cloud of the handle part.

Point clouds obtained with different Outlier Removal values (all other conditions identical):

Off Weak

outliers removal off

outliers removal weak

Normal

Strong

outliers removal normal

outliers removal strong

Noise Removal

Description

This parameter removes the noise in the point cloud, thus reducing the impact on the precision and accuracy of subsequent calculation. Noise is the scattered points close to the object surface.

Visibility

Expert, Guru

Values

  • Off

  • Weak

  • Normal

  • Strong

Instructions

  • Higher intensity removes more noise but may also remove some object features. Lower intensity removes less noise but also preserves the object features better.

  • Higher intensity results in slower calculation, and lower intensity results in faster calculation.

Point clouds obtained with different Noise Removal values (all other conditions identical):

Off Weak

noises removal off

noises removal weak

Normal

Strong

noises removal normal

noises removal strong
If this function removes the needed object features, please reduce the intensity. However, more noise will be retained.

Edge Preservation

Description

This parameter preserves the sharpness of object edges during surface smoothing.

Visibility

Guru

Values

  • Sharp: Preserves the sharpness of object edges as much as possible. However, the effect of surface smoothing will be reduced.

  • Normal: Balances between edge preservation and surface smoothing.

  • Smooth: Does not preserve the edges. The object surface will be well smoothed, but the object edges will be distorted.

Instructions

Adjust based on your needs for preserving the edges of workpieces.

Stripe Contrast Threshold

Description

This parameter is used to remove noise in the point cloud. Adjust this parameter if the level of noise in the point cloud is still high after adjusting Outlier Removal and Noise Removal.

Visibility

Beginner, Expert, Guru

Values

1 to 100

Instructions

  • Greater value of this parameter results in more points being removed. Smaller value results in fewer points being removed.

  • Increase the value of this parameter can remove the noise in the point cloud, but may also cause the points of dark objects to be lost.

Point clouds obtained with different Stripe Contrast Threshold values (all other conditions identical):

3 15 30

signal contrast threshold 3

signal contrast threshold 15

signal contrast threshold 30

Depth Range

Description

Set a Z-direction ROI for the depth map and point cloud. Depth Range can be set within the camera working distance, and the data outside the range is removed from the depth map and point cloud.

Visibility

Beginner, Expert, Guru

Values

  • Lower Limit: 1 to 4000 mm

  • Upper Limit: 1 to 5000 mm

Instructions

  • Depth Range should be adjusted to an appropriate range to ensure the completeness of depth data in the depth map and point cloud. Too large a range may introduce interference, and too small a range may remove the data of key features.

  • For detailed instructions, please refer to Set Depth Range below.

The same point cloud with different Depth Range values (all other conditions identical):

Too large Appropriate Too small

depth large

depth moderate

depth small

depth large 1

depth moderate 1

depth small 1

Set Depth Range

Follow these steps to adjust Depth Range:

  1. Click Edit on the right of Depth Range to open the Set Depth Range window.

  2. Click Refresh point cloud at the top of the right panel to obtain the newest point cloud.

  3. Adjust the position of the point cloud: adjust the position until you can see the two gray rectangles that represents the upper and lower limits of Depth Range.

    set depth range

  4. Adjust the depth range: drag the handles on the slider to adjust Depth Range roughly. Then, enter specific values on the right to fine-tune Depth Range.

    Judge if the depth range is appropriate: All needed object features are located between the two gray rectangles, and most noise and outliers are located outside.

  5. Click Save to save the depth range settings.

  • Click Recommended range to adjust Depth Range to the recommended working distance of the current camera.

  • Click Reset to restore Depth Range to the last saved values.

ROI

Description

Set the ROI on the XY plane for the depth map and point cloud. Points outside the selected region are removed.

Visibility

Beginner, Expert, Guru

Values

N/A

Instructions

For detailed instructions, refer to Set ROI below.

Set ROI

  1. Click Edit on the right of ROI to open the Set ROI window.

  2. Select and adjust the ROI on the left. Drag the selection box to move it. Drag the handles on the box to adjust the box size.

  3. Click Apply to apply the set ROI.

    quick setroi

    • Click Clear to remove the current ROI.

    • For the DEEP-GL (V4) and LSR (V4) series, the 2D image (depth source) is displayed in this window. If the image is too dark or too bright, please adjust 2D Image (Depth Source) Exposure Mode.

  4. Acquire data again and view the depth map or point cloud to check the result of setting the ROI.

    quick setroi 1

We Value Your Privacy

We use cookies to provide you with the best possible experience on our website. By continuing to use the site, you acknowledge that you agree to the use of cookies. If you decline, a single cookie will be used to ensure you're not tracked or remembered when you visit this website.

  • Preference
  • I accept
  • I don't accept