Welcome to PIMFP !
PIMFP is a web tool for the prediction of IMF percentage in live lean pigs using backfat ultrasound images. For now, PIMFP model was trained by a large-scale dataset consisting of IMF percentage and 4,552 ultrasound images of 945 pigs. The training data will constantly increase in the future.
Please use our service after signing up.
1. Select ultrasound equipments:
Users need to first select the ultrasound equipment (click on "Ultrasound equipment"), and currently, we support two ultrasound equipments: EXPRO and Unicorn Vet.
* For the EXPRO device, this website supports several common image export formats with the following pixel specifications:
① 1079*680 pixels; ② 1138*854 pixels; ③ 1215*765 pixels; ④ 580*480 pixels
* For the Unicorn Vet device, this website supports several common image export formats with the following pixel specifications:
① 940*720 pixels；② 880*688 pixels.
You can download example images from the manual of PIMFP.
Please note that images that do not meet the above pixel specifications will not be allowed for upload. If you have images exported from other ultrasound equipments or have any questions about the image formats mentioned above, please contact us at zhenliu@cau.edu.cn. We welcome users with other ultrasound equipments to contact to us, and in the future, this website will support more ultrasound devices and image formats.
2. Upload image:
After selecting the backfat scanner device, users can click "select image" to upload backfat ultrasound images.
Once you have completed the registration process, you will be considered a regular user by default.
* For regular users, you are allowed to upload up to 100 images per day. If the cumulative number of predicted images exceeds the limit for the day, you will not be able to upload any more images until the next day. If the regular user limit does not meet your needs, you can contact zhenliu@cau.edu.cn to upgrade for VIP user.
* For VIP users, there is no limit on the cumulative number of uploads per day. There is a maximum limit of 500 images per single upload.
3. Predict IMF percentage:
Once all the preview images on the website are displayed, users can click on "Predict" to predict the IMF percentage. If you have uploaded a large number of images at once, it may take some time for the preview images to load. Please be patient and wait, or you can upload them in batches. The prediction results will be sent to your email.
This website does not collect your personal information. During the use of our website, we do not record your IP address, browser type, referral address, or any other technical data. The images uploaded by users will be temporarily stored on the website's server, and immediately deleted after the prediction results are sent to the user's email. This website does not download, browse, or copy the images uploaded by users.
Section 1: User Commitments
During the use of this website and its services, users are prohibited from engaging in any activities that violate Chinese laws and regulations or social ethics. Users are expected to adhere to the following commitments:
1. The uploaded, posted, and shared image content must comply with Chinese laws, regulations, and social ethics.
2. The uploaded, posted, and shared image content must not infringe upon the legitimate rights of others.
3. Users must not interfere with, harm, or infringe upon the lawful rights and interests of this website.
4. Users must comply with the agreements, guidelines, and regulations of this website and its related online services.
5. This website reserves the right to delete content that violates the above commitments.
Section 2: Disclaimer
This website serves as a network platform and storage space for users to upload, post, and share image content. The uploaded, posted, and shared image content by users solely represents their own views and opinions, and does not reflect the views and opinions of this website. As the uploader of image content, users are responsible for their uploaded content. This website shall not be held liable for any disputes or losses arising from the published image content. The uploader of the image content shall bear all direct or indirect (joint and several) legal liabilities. This website assumes no legal responsibility.
The image content uploaded, posted, and shared by users on this website is generated through their individual actions. Any disputes or losses arising from the profit-making or non-profit-making activities conducted by individuals, companies, or other entities using the aforementioned content that infringe upon the intellectual property or other legitimate rights of others shall be borne solely by the individual, company, or other entity. This website shall not be held liable for any legal responsibilities.
In the event that the image content uploaded by users on this website infringes upon the intellectual property or other legitimate rights of others, or upon receiving complaints or reports from relevant rights holders, this website has the right to delete the infringing content or take other restrictive measures. This website also reserves the right to transfer the case to relevant administrative authorities for handling. This website shall not be held liable for any losses or other legal liabilities incurred as a result.
If you have any questions, please feel free to contact us: zhenliu@cau.edu.cn. We wish you have a pleasant experience while using our service!
1. Example ultrasound image.
2. If you are a Chinese user, you can download the manual in Chinese.
Liu Z, Du H, Lao F D, et al. PIMFP: An accurate tool for the prediction of intramuscular fat percentage in live pigs using ultrasound images based on deep learning[J]. Computers and Electronics in Agriculture, 2024, 217: 108552.
