Study of color and textural feature variation of carp meat using image processing
Fish quality is affected by terms of handling, maintenance and storage time. These terms make chemical changes in fish and accelerate the deterioration its of tissue and make it dangerous for human body. There are several methods use for assessment of fish freshness, most of them are costly and destructive., Therefore, in this paper a non-destructive machine vision system based on gill and eye color and textural features is proposed. . Accordingly, after segmentation of region of interest in the images (eyes and gills), the color and textural properties of the images were extracted and the most suitable ones were selected using Fisher's selection algorithm and QDA and LDA classification methods were applied. For the QDA classifier , the V_HSV (extracted from the gills), the energy and the contrast (extracted from the fish's eye) and for the LDA classifier, the energy (extracted from the eye), the contrast )extraction from the eye) ,V_HSV (extracted from the gills) ,homogeneity (extracted from the eye) and H_HSV (extracted from the gills) were extracted. The classification accuracy for QDA and LDA were 93% and 96%, respectively.
پرداخت حق اشتراک به معنای پذیرش "شرایط خدمات" پایگاه مگیران از سوی شماست.
اگر عضو مگیران هستید:
اگر مقاله ای از شما در مگیران نمایه شده، برای استفاده از اعتبار اهدایی سامانه نویسندگان با ایمیل منتشرشده ثبت نام کنید. ثبت نام
- حق عضویت دریافتی صرف حمایت از نشریات عضو و نگهداری، تکمیل و توسعه مگیران میشود.
- پرداخت حق اشتراک و دانلود مقالات اجازه بازنشر آن در سایر رسانههای چاپی و دیجیتال را به کاربر نمیدهد.