To increase the field of view (FoV) of the Raspberry Pi Camera (IMX219), one could you an adapter to convert the camera into a wide angle fisheye lens. In this post, we’ll be looking at the Entaiya Wide angle fish eye lens, which have a large Horizontal FoV of 220° !
Here are the details of the lens (mounting bracket) adapter:
|Dimensions||35 x 21.7 x 24mm|
The kit consists of 6 packs of parts, which includes the mounting plate, the adapter, the spacer, the screws, the rubber grip and the lens.
Note that the camera is not included in the kit!
Disclaimer: This guide serves as a reference for you to install the Fish eye adapter, please proceed on with caution only if you are agreeable to the fact that we would not be responsible for any issues or damage that may occur whilst following the steps.
Unlatch the sensor
First, unlatch the camera from the adapter. The camera would not come of yet as there is a double side tape/glue sticking to the back of the camera sensor. Carefully pull the camera sensor away from the green body plate.
After the camera sensor is separated from the body plate, remove the glue behind the camera sensor. (The glue could not be reused as the sensor would not be able to fit properly on the kit’s body plate, which is designed to fit perfectly without the tape/glue.)
Body plate attachment
After that, screw in the body plates as shown in the instruction sheet. Attach a thin layer of blue-tack/glue behind the camera sensor before reattaching to the body plate. (Ensure that the blue-tack/glue does not cause the sensor to bulge out too much, as the camera sensor is supposed to be flat on the body plate.)
After the camera sensor has been re-attached, assemble the clamp as shown in the instruction sheet. This would be used to remove the original lens on the camera. Slowly tighten the clamp with the provided Allen key, ensuring that only the lens mount is clamped. (Tighten enough such that the clamp does not fall off from the camera lens, do not over-tighten!)
Please finish reading the entire section before proceeding on, as this is the trickiest step. Please proceed with caution & be careful when removing the lens!
With the clamp on, pull the clamp away from the attachment point using the corners as the pivot point. If the camera sensor shifts completely or the lens does not come out, DO NOT FORCE IT OUT! This means the clamp is clamping onto the camera sensor too. Loosen the clamp and re-clamp the lens.
If the clamp is correctly aligned with the camera lens, the clamp would be able to remove the lens easily by pulling it gently.
After the camera lens have been removed, place the rectangular rubber behind the new adapter and attach it on the body plate. Screw in the 2 screws onto the adapter, which is located on the center sides of the adapter. Next, insert the waver spacer into the lens before screwing the lens into the adapter. And viola! The adapter has been installed.
Adjusting the focus
After that, attach the the ribbon cable back to the relevant development board/SBC (single board computer) to get the image feed. The image would be very blur at first, adjust the camera focus accordingly by screwing it clockwise/ant-clockwise.
Mostly minor issues
- The alignment of the camera can be tricky as you have to tighten the screws without shifting the lens: any slight movement will cause the sensor to be off center and the image would not be centered.
- There are minor scratches on the lens, which may affect the image quality.
- The focus may not be uniform: when the center is focused, the sides are slightly blurred, vice versa.
- The internal reflection of the barrel (beside the spherical image), although it could be masked through image processing like OpenCV. (If you want to mask the image sides in ROS, do checkout masker_util, a ROS package that I have created just to do just that! It uses dynamic reconfigure so that you could specify the location of the mask.)
The adapter kit is a good way to convert the normal (FoV) camera into a wide angle fish-eye lens, which captures quite a lot of the scene into an image. This is useful in capturing a huge field of view, with various applications such as capturing large architecture, creating panorama images, VR or even creating 360 degrees images or video.