I'm working in a RGB image classification algorythm for blueberry ripeness stages identification in which I must collect photos for the algorythm traning data set. Is it valid taking the photos using a smartphone camera with enought resolution (eg.: Xiami Redmi Note 10, etc.), considering this research is intended to be published on a scientific journal? I've read some papers refering usage of GoPro cameras for collecting image-data sets. I have compared the image quallity between my smartphone and GoPro pics and the first one are quite better.
Thanks for the answers!
In principle I can't see an issue using a smartphone camera assuming it aligns with the context of your research. Just list the device details in the paper along with the findings.
As David mentioned above, there is no restriction on the type of cameras / resolution one can use in research. In the context of smart farming, many of the recent research works are using images captured by a flying drone from a certain altitude.
Dear Héctor Andrés Ramírez
These articles might be helpful, have a look:
1. https://link.springer.com/article/10.1007/s00216-021-03184-z
2. https://www.spiedigitallibrary.org/journals/journal-of-biomedical-optics/volume-26/issue-04/040902/Smartphone-based-imaging-systems-for-medical-applications--a-critical/10.1117/1.JBO.26.4.040902.full
Kind Regards
Qamar Ul Islam
If you use Python software and yolo, maybe I can help you. First, you have to get the network IP address and port of your smartphone, and then type the code: python detect.py --source http://admin:admin@****:##, **** is your smart The phone's network IP address, ## is the port.
Even low-end smartphones can take pictures with sufficient resolution for most of the analysis. I invite you to visit our paper. We also tested the impact of the image resolution on the accuracy of disease severity measurements using RGB images. Preprint Measuring plant disease severity in R: introducing and evalu...
Héctor Andrés Ramírez
...and here I cite my MSc student who did a study on the effect of resolution. Her thesis is shown below with a link to the full thesis:
Devagiri, V. M. (2017). Splicing Forgery Detection and the Impact of Image Resolution (Dissertation). Retrieved from:
http://urn.kb.se/resolve?urn=urn:nbn:se:bth-14060
It would be acceptable to use a smartphone of that quality for the capture of the images. Also, if you have access to a GoPro you could analyse the RGB values of the GoPro and the same image with the smartphone to validate the smartphone values.
Many researchers use GoPro not only for the image quality, but also for the high frame rate. Several years ago, GoPro was one of the cheapest cameras with 120 fps at 720p resolution and higher. Today, the smartphone can reach these resolutions. Another feature of GoPro is the standard tripod connector and for smartphones you must buy one.
- Badges
- Science topic
- Similar topics
- Phytochemistry
- Extracts