Text
PERBAIKAN CITRA JANTUNG JANIN MENGGUNAKAN METODE LOW-LIGHT CONVOLUTIONAL NEURAL NETWORK (LLCNN)
In Computer Vision processing has some problems which one quite often happens especially in the medical field is lack of brightness or contrast of an image. High quality images are required to get better results. Image enhancement process is process which have purpose to doing enhancement quality for some image from have low quality image to good quality image with in a way that is doing image improvisation that have low quality image before to gain good quality image. This fetal heart image enhancement process can use a deep learning method is LLCNN. And next will be doing verification such as doing 2 class classification with normal and abnormal class using some transfer learning for comparison. The research result for LLCNN method get average value from 6 medical image with score 18.6% Mean Squared Error (MSE), 96.03% Structural Similarity Index Measure (SSIM), and 30.86 dB Peak Signal Noise Ratio (PSNR). And for classification result the best result on DenseNet121 transfer learning with accuracy result before using enhancement model on validation result is 97% and unseen result is 93%, and after using enhancement model on validation result is 100% and unseen result is 100%.
Inventory Code | Barcode | Call Number | Location | Status |
---|---|---|---|---|
2207003451 | T77273 | T772732022 | Central Library (Referens) | Available but not for loan - Not for Loan |
No other version available