Blood cell segmentation from microscopic blood images

Abstract

Blood cell segmentation is a critical innovation for automatic differential blood counting, classification and analysis in clinical examination. In color blood cell images segmentation and recognition are two essential issues in the field of biomedical cell morphology. This paper approaches methods to segment the blood cells from microscopic thin blood images. This data is the premise to perform higher level tasks for example, automatic differential blood counting, detection of different diseases such as Malaria, Babesia, Chagas disease, Anemia, Leukemia etc. A noteworthy necessity of an automated, real-time, computer vision-based cell segmentation system is an efficient method for segmenting different blood component such as Red Blood Cells (RBCs), White Blood Cells (WBCs), Platelets from input images for blood count as well as to detect the parasites present in blood cells. Input images are captured by connecting digital camera to microscope. Captured images are enhanced and segmented using K-mean clustering as well as global threshold. Overlapping cells are separated using Sobel edge detector and Watershed transform.

Overlapping White Blood Cells Detection Based on Watershed Transform and Circle Fitting

“…To evaluate the performance, the proposed and existing DE [8], WS [16], CHT [17], CRT [18] schemes have Noise Type Technique DR (%) FAR (%) None WS [16] 82.23 % 12.00 % CHT [17] 75.09 % 16.66 % DE [8] 70.10 % 13 % CRT [18] 78.02 % 14.12 % Proposed 97.10 % 7.80 % Salt & Pepper (5 %)…”

“…WS [16] 79.00 % 10.22 % CHT [17] 68.56 % 16.78 % DE [8] 68.00 % 11.46 % CRT [18] 73 been simulated on ALL-IDB [9] and ASH image bank [10] databases. The databases contain a total of 384 WBC images including some overlapping cases.…”

“…2(c) and 2(f) detects almost all the WBCs. Table 1 shows the quantitative comparison for the existing DE [8], WS [16], CHT [17], CRT [18] and proposed schemes. The comparison with CRT [18] might be unfair since their segmentation mainly focuses on RBCs rather than WBCs.…”

“…The comparison with CRT [18] might be unfair since their segmentation mainly focuses on RBCs rather than WBCs. It can be viewed that the proposed approach provides better DR and FAR as compared with DE [8], WS [16], CRT [18] and CHT [17] even in presence of noise.…”

“…Automatic counting and classification of WBCs based upon morphological features in [15] can not detect irregular and overlapped WBCs accurately. In [16], edge detection and watershed segmentation (WS) have been utilized to separate overlapping WBCs but mostly the separation of overlapping WBCs suffers from over-segmentation even after the application of post-processing steps. The post-processing steps used include erosion and dilation with a structuring element of size 2, which fails to provide accurate segmentation in case of maximum overlapped WBCs.…”