Further, some areas of science still use the conventional approach of solving the problem, and sericulture is one among them. The sericulture industry involves the art and science of host plant cultivation as well as silkworm rearing to produce natural silk products. Silk is the queen of textiles and globally India is the second-largest producer of four different types of silk. Thus, sericulture serves as the base for economic, social, scientific, political, and intellectual advancements [4]. The fecundity (number of eggs laid by fertilized female silk moth), hatching percentage (silkworm birth rate), survival percentage (disease and environment tolerant), and silk productivity are a few economic traits (parameters) on which entire silk industry thrives. Manual counting of eggs is in vogue to quantify fecundity and hatching percentage parameters. Many automatic methods (image processing and new hardware design) have been attempted with lower accuracy [5]. A new approach of automatic counting and classifying eggs is described in this paper to quantify fecundity and hatching percentage accurately which provides required rearing information to harvest successful silk cocoon crops.
The chapter describes a few conventional approaches and their drawbacks and, further, introduce the CNN approach adopted in this paper and to explain the specifications of each model trained to surpass the results provided by other image processing techniques.
2.2 Conventional Silkworm Egg Detection Approaches
Manual counting of silkworm egg is in practice in countries like India, China, Thailand, and other Asian countries [6]. The silkworm eggs are small-sized [5], approximately 2 to 3 mm in diameter, densely populated in small clusters. Hence, the manual counting process will be tediously associated with prolonged time and is susceptible to human error. The inconsistency in determining the fecundity and hatching percentage impacts the overall cocoon crop performance and Silk productivity.
Many techniques have been implemented to measure the quantity of egg laid, such as designing hardware [6, 7], and using image processing techniques. The primary focus in previously published papers was to segment the silkworm egg from the background using different image processing techniques such as low contrast image setting [8], contrast enhancement followed by image morphological operations [7], image patch centroid analysis [6], image channel conversion from RGB to HSV to identify the region of interest (ROI) [9], using Gaussian mixture model [10], and using Hough transforms (blob analysis) [11, 12]. The accuracy achieved in these techniques completely depends on the consistency of the experimental conditions such as the color of the sheet on which the silkworm eggs were laid, the size of the eggs, and uniform illumination while capturing a digital image of eggs. By altering any one of these parameters, the results vary drastically, and hence designing an image processing algorithm for every possible scenario becomes laborious. Also, the method used in these techniques to capture digital data of the eggs was to use digital cameras, operated manually without any preset illumination parameter and hence resulting in poor accuracy.
2.3 Proposed Method
Two of the main parameters that vary during capturing digital data for image processing are the size of the silkworm egg and uniform illumination spread across the image. Firstly, since the image processing (including blob analysis) algorithms are designed to identify a particular egg size or range of egg sizes, exceeding this limit causes error in the final result. Since no constant distance is set between the egg sheet and camera, in any of the earlier papers, the pixel size of captured eggs varies which causes the problem to the image processing algorithm. Also, the irregular distribution of illumination over ROI causes the digital cameras to record the data slightly in a different way, which may over saturate or under saturate the ROI. The image processing algorithms such as contrast stretch and histogram equalization perform well on the limited scenario and do not provide complete confidence to enhance low-quality data.
To overcome these issues, a constant illumination light source with a fixed distance between camera and egg sheets of a paper scanner is used to capture the digital data of the silkworm egg sheets. Since the distance between the camera array of the paper scanner is fixed, the egg size can be approximated to stay within a specific range, i.e., around 28 to 36 pixels in diameter in our experiment. However, not all manufacturers of paper scanner follow strict dimensions while designing, hence the silkworm eggs scanned with different scanner results are found to be different. For example, the eggs scanned with Canon® scanner have a diameter of 28 to 32 pixels under, while 36 to 40 pixels with Hewlett-Packard® (HP) scanners for the same resolution and dots per inch (dpi).
Figure 2.1 Adding a key marker on the silkworm egg sheet.
Also, by changing the scanner parameters such as resolution and dpi, the resulting egg diameter is of different pixels size for the same scanner. Hence, a key marker is printed on the egg sheet before it is scanned to capture the details in a digital format. The dimension of the key marker is 100 × 100 pixels (10 × 10 mm) which are considered as a standard dimension in our experiment. Let R(hxw) be the standard resolution required by the image processing algorithm, while R′(h’xw’) be the resolution at which the egg sheet is scanned and R″ is the resulting resolution of the image. Also, let (Dx, Dy) be the standard dimensions of the key marker, while
(2.2)
2.3.1 Model Architecture
To overcome the problem with conventional image processing algorithms as stated in Section 2.2, machine learning (ML) techniques were employed to segment the eggs from the background (egg sheet). The features to be considered for segmentation such as color (grayscale pixel values), the diameter of eggs was collected manually using image processing and feature engineering,